N-myristoyltransferase

Molecular and Cellular Biochemistry 204: 135–155, 2000. © 2000 Kluwer Academic Publishers. Printed in the Netherlands.

135

N-myristoyltransferase Raju V.S. Rajala,1 Raju S.S. Datla,2 Terence N. Moyana,1 Rakesh Kakkar,1 Svein A. Carlsen1 and Rajendra K. Sharma1 1

Department of Pathology and Saskatoon Cancer Centre, College of Medicine, Royal University Hospital, University of Saskatchewan, Saskatoon, Saskatchewan; 2Plant Biotechnology Institute, National Research Council, Saskatoon, Saskatchewan, Canada Received 13 May 1999; accepted 17 September 1999

Abstract Myristoylation refers to the co-translational addition of a myristoyl group to an amino-terminal glycine residue of a protein by an ubiquitously distributed enzyme myristoyl-CoA:protein N-myristoyltransferase (NMT, EC 2.3.1.97). This review describes the basic enzymology, molecular cloning and regulation of NMT activity in various pathophysiological processes such as colon cancer and diabetes. (Mol Cell Biochem 204: 135–155, 2000) Key words: myristoyltransferase, myristoyl-CoA, lipid modification, purification, expression

Introduction Protein modification is an important aspect of gene regulation. Translation of the genetic code is only one step in the sequence of events that determine the steady state level of each of the multitude of mature proteins found in the cells. Most proteins are also subjected to a wide variety of co- and post-translational modifications. The structural alterations resulting from these modifications induce or control biological activity, direct intracellular and/or extracellular translocation, and effect stability and turnover. The structure of many proteins in their mature functional state includes various kinds of covalent modifications that have been added onto the protein’s polypeptide backbone. The common covalent modifications are by addition of oligosaccharides, phosphate, acetyl groups, formyl groups or nucleotides. Their addition to the polypeptides occurs both during nascent chain biosynthesis and later as the protein becomes integrated into cellular metabolic activity. It is clear that a defined sequence of amino acids is the primary determining factor in controlling such modification. How these modifications affect cellular function is known in only a few cases. Covalent attachment of lipids are important components of a diverse array of cellular proteins. The proteins acylation

by long chain fatty acid or ‘sticky fingers’ appears to be specific for linkage and promotes the binding of these proteins to the inner face of the cellular plasma membrane. Three common lipid modifications have been identified and are important in cellular regulation: myristoylation, Spalmitoylation and prenylation. A fourth modification involves addition of a glycosylphosphatidylinositol (GPI) moiety to proteins so that they can be anchored to the plasma membrane [1]. GPI anchoring and prenylation will not be discussed here but have been reviewed recently [1]. In general, palmitate is linked to newly-synthesized proteins via ester or thioester linkages to serine (or threonine) and cysteine, respectively. This modification occurs posttranslationally, probably in the cis Golgi complex or in the transitional elements of the endoplasmic reticulum [2–6]. Whereas myristate becomes covalently bound to both soluble and membrane proteins via an amide linkage to an aminoterminal glycine residue of the known N-myristoyl proteins [7–11]. However, in a few cases this N-terminal glycine specificity for N-myristoylation is not observed [12]. In addition, Muszbek and Laposata [13] have demonstrated that in platelet proteins, myristate is linked to proteins predominantly via thioester bonds, as is palmitate, and that the covalent binding of these two saturated fatty acids to proteins

Address for offprints: R.K. Sharma, Department of Pathology, Saskatoon Cancer Centre, 209, 20 Campus Drive, University of Saskatchewan, Saskatoon, Saskatchewan, S7N 4H4, Canada

136 involves the same mechanism. Palmitate is a common metabolite in eukaryotic cells, but myristate is present in only small amounts (less than stearic acid), suggestive of a special role for myristoylation which would not be served by the addition of any other lipid. This high degree of specificity of palmitate and myristate for ester and amide linkages, respectively, suggests the possible existence of distinct classes of protein acyltransferases. The myristoylation of proteins has received considerable attention recently, due to observations that many proteins involved in cellular regulation and/or signal transduction are myristoylated including the catalytic subunit of cAMPdependent protein kinase [7], various oncoproteins (including pp60src, pp56lck) [11, 14–16], the β subunit of calmodulindependent protein phosphatase (calcineurin) [8], the myristoylated alanine rich C kinase substrate (MARCKS) [17], the α subunit of several G proteins [18], and several ARF proteins involved in ADP ribosylation [19]. Since myristoylated proteins are localized in both cytosolic (i.e. catalytic subunit of cAMP-dependent protein kinase and β subunit of calcineurin) and membrane associated cellular compartments (i.e. endoplasmic reticulum for cytochromeb5 reductase, plasma membrane forpp60 src), it has been suggested that myristoyl addition plays a more complex role than simply increasing the hydrophobicity of the proteins for membrane anchoring. Recently, it has been shown that dephosphorylation of the catalytic subunit of myristoylated and non-myristoylated cAMP-dependent protein kinase at Thr197 by cellular protein phosphatase and protein phosphatase2A (PP-2A) indicated that the myristoylated C subunit was more resistant to dephosphorylation than the non-myristoylated enzyme [20]. Myristoylation has been shown to be necessary for the transforming potential [11] and receptor recognition of pp60 v-src [21, 22]. Myristoylation is also required for activation of pp60c-src tyrosine kinase activity during mitosis [23]. N-myristoyl CoA: protein N-myristoyltransferase (NMT, EC EC 2.3.1.97) is the enzyme that catalyzes the covalent transfer of myristic acid (a 14-carbon fatty acid) to the Nterminal glycine residue of a protein substrate (for reviews see refs [12, 19, 24, 25]). Mutations that convert an N-terminal glycine to alanine completely abolishes myristoylation [26– 29]. Also, chemical analysis of the fatty acid that is linked to the amino-terminal glycine yields only myristic acid [8–10, 30]. Stuck et al. [31] have reported that acylation of mitochondrial proteins from rat liver is reversible and very rapid. In addition, several reports have indicated that myristoylation may also occur post-translationally [32–34]. A reversible N-myristoyl linkage has been described in Dictyostelium discoideium [35]. One possible explanation for the existence of a pool of non-myristoylated protein in the cell is the existence of a demyristoylase, as has been observed in brain synaptosomes [36]. The purpose of this review is to

summarize some of the most significant advances in research on NMT which have been carried out in our laboratory during the past several years.

Results N-myristoyltransferase(s) Towler and Glaser [37, 38] were the first to demonstrate the existence of NMT in Saccharomyces cerevisiae by using the N-terminal sequence (Gly-Asn-Ala-Ala-Ala-Ala-Arg-Arg) of the catalytic subunit of cAMP-dependent protein kinase, a myristoylated protein in the presence of myristoyl-CoA and ATP. In addition, Towler et al. [38] have shown the initial purification and characterization of NMT and reported that NMT has an absolute requirement for N-terminal glycine and myristoyl-CoA [37, 38]. The S. cerevisiae NMT was the first enzyme which was purified and characterized extensively [39]. The purified NMT from S. cerevisiae is a monomer with an apparent molecular mass of 55 kDa [37, 38]. Comparison of S. cerevisiae and rat liver NMT indicated overlapping, yet distinct peptide substrate specificities [40]. Furthermore, this study demonstrated that synthetic peptides derived from sequences of known myristoylated proteins found in higher eukaryotes were substrates for yeast, rat liver and wheat germ cell NMT, however, there were differences in peptide substrate specificities [40, 41]. For example, the peptides corresponding to the N-terminal sequence of transducin-α, Giα and SYN kinase were substrates for rat liver NMT but not yeast NMT [40]. These results suggest that either the enzyme from yeast or the multiple forms of NMT in rat liver possesses distinct substrate specificities. Originally, our laboratory has reported that bovine brain possesses multiple forms of NMT [42]. When the bovine brain high speed supernatant fraction was applied onto a DEAE-Sepharose CL6B column, four distinct enzyme activity peaks were obtained and were designated as peaks I, II, III and IV according to their order of elution. Proteolytic inhibitors were included in the isolation buffer as well as in the purification buffers. The results from gel filtration column chromatography as well as the kinetic parameters further suggest that bovine brain may contain multiple forms of NMT [42]. Since the brain is a highly heterogeneous organ containing many different cell types, it is possible that the different forms of NMT observed from DEAE ion-exchange chromatography represent different isozymes from different regions of the brain. Furthermore, it has been observed that NMT activity was not uniformly distributed in different regions of rat brain [43]. To further substantiate that bovine brain contains multiple forms of NMT, Western blot analysis was carried out using polyclonal antibodies against NMT [44, 45]. Our recent results demonstrated that multiple forms of

137 NMT, in addition to bovine brain, were also present in skeletal muscle and lung, however, a single form was observed in pancreas, liver, spleen, and heart (Fig. 1) [45]. Therefore, these observations further suggest that NMT exists in different isozymic forms. The apparent molecular mass was dependent on the protein concentration of the sample applied to the gel filtration column, since high protein concentration resulted in a high apparent molecular mass while lower protein concentrations resulted in lower apparent molecular masses [42, 46]. These results are consistent with the theory that NMT is a ‘sticky’ protein aggregating either

Fig. 1. (A) Northern blot analysis of total RNA in various bovine tissues. Ten microgram of total RNA was electrophoresed on 1.2% agarose gel containing 6.7% formaldehyde, followed by blotting onto a nitrocellulose membrane. Hybridization was carried out using a sNMT cDNA probe; (B) Western blot analysis of various bovine tissues with anti-peptide antibody. Sixty micrograms of protein was subjected to SDS-polyacrylamide gel electrophoresis and immunoblotted with affinity purified anti-peptide antibodies. C. Bovine tissue enzyme activities were determined in cytosol fractions in the presence of cAMP-dependent protein kinase peptide substrate. For experimental conditions see Raju et al. [45].

with itself or with other proteins; the lowest molecular mass (51 kDa) might represent a single catalytic subunit and the higher molecular masses might represent homogenic or heterogenic multimers (possibly dimers and tetramers). The NMT ‘aggregation’ is probably due to specific protein interactions, dissociated by myristoyl-CoA (200 mM) but not by high salt concentration (1 M NaCl) [42, 46]. McIlhinney et al. [47] also identified two forms of NMT in bovine brain cortex in the presence of myristoyl-CoA on gel filtration. Bovine cortex 66 kDa NMT has been purified and partially sequenced [47]. Glover and Felsted [48] reported that bovine brain NMT exists in vitro as two multisubunit complexes which can be interconverted (390 kDa [25%], 126 kDa [75%] and 50 kDa [1–2%]) by manipulation of ionic strength and/ or by treatment with an N-myristoyl-peptide. Five isoforms of NMT have also been observed in the murine leukemia cell line L1210 [49]. One of the isoforms was purified, which had an apparent molecular mass of 67.5 kDa [49]. A 63 kDa human HeLa cell NMT has been characterized immunologically [50]. Progress in biochemical characterization of mammalian NMT has been slow because of difficulties in purifying significant amounts from tissue extracts. A human NMT (hNMT) has been previously obtained by functional complementation in yeast, and the expressed protein has been characterized [51], however several laboratories suggest that other isoforms may exist in mammalian cells [40, 44–48]. To facilitate the purification of NMT, bovine spleen was chosen because it is available in large quantities and is a rich source of NMT activity. In addition, pp60src is present in abundant quantities in spleen [52]. Using the combination of ammonium sulphate precipitation, chromatography on SP-Sepharose fast flow, phenyl-Sepharose CL-4B, DEAE-Sepharose CL-6B, and Superose 12 (HR/30) gel filtration fast protein liquid chromatography, the enzyme was purified close to homogeneity with a high yield [53]. Under native conditions, the enzyme exhibited an apparent molecular mass of 58 kDa, whereas under denaturing conditions the enzyme represented an apparent molecular mass of 50 kDa, suggesting that spleen NMT (sNMT) is a monomeric protein [53]. In a survey of various rat tissues, McIlhinney and McGlone [43] reported a high level of NMT activity in brain and gut. Independently King and Sharma [32] also observed high NMT activity in rat and bovine brain tissues. We also investigated the distribution of NMT activity in the intestine of rabbits and determined that the highly proliferative epithelial mucosa contained higher NMT activity than the smooth muscle [54]. Regional differences in NMT activity throughout the rabbit intestinal tract were observed as NMT activity decreased from the small intestine to the descending colon [54]. Descending colon NMT has an apparent molecular mass of 78 kDa [54]. The physiological significance of different levels of mucosal NMT activity in various regions of the intestine is unknown at this time. There have been

138 several myristoylated proteins identified in intestinal cells [55–58]. Normal intestinal epithelial cells express high levels of pp60c-src and pp60c-yes [59, 60]. Intestinal specific annexin is a myristoylated 35 kDa protein localized in the membrane of HT-29 cells derived from human colon carcinoma [55]. Although the specific role of this protein is not known, other annexins have been implicated in signal transduction and cell growth and regulation [56]. Cell adhesion regulator (CAR), a gene implicated in the suppression of tumor invasion and cloned from a colonic cell line, encodes a 142 amino acid protein which has an N-terminal myristoylation site and a Cterminal tyrosine phosphorylation site [57]. Heparin sulfate proteoglycan from WiDr colon carcinoma cells had myristate bound covalently through a hydroxylamine and alkali resistant bond [58]. Enterokinase, a protease of the intestinal brush boarder, has been sequenced and found to contain the known sequence requirement for N-myristoylation [61]. Since the intestinal mucosa represents a population of rapidly proliferating cells, this tissue may be a useful model to study the role of NMT in normal cell physiology.

Kinetic parameters The kinetic properties of NMT have been examined by various laboratories from a variety of tissues and animal species using either crude, partially or highly purified NMTs [40, 46, 53, 54, 62–67]. The kinetic parameters are given in Table 1. In general, NMT isozymes have a higher affinity towards pp60src peptide [40, 46, 53, 54, 62–67].

Molecular cloning and biochemical characterization of sNMT The availability of cDNA clones of this enzyme will assist in identification of essential motifs involved in enzyme activity. We have isolated the cDNA encoding bovine sNMT from a λ gt11 bovine spleen cDNA library employing a hNMT cDNA probe [44]. The isolated insert contained 1703 bp with an open reading frame of 1248 bp [45]. This open reading frame is flanked by 136 bp of 5′ untranslated sequence and 316 bp of 3′ untranslated sequence. The ATG initiation codon is found at position 137. The single long open reading frame of bovine spleen cDNA specifies a protein of 416 amino acids with a predicted molecular mass of 46,686 Da. The open reading frame terminates with TAA stop codon at position 1385 [45]. The sequence generated from the spleen cDNA clone was compared with that of human cDNA and the analysis showed a high degree of conservation at both the nucleotide and protein levels [44]. Within the coding region there is 92.4% identity with the hNMT at the nucleotide level. Sixty nine

percent identity is observed in the 5′ untranslated sequence and 85% in the 3′ untranslated sequence. Multiple sequence alignment revealed that the predicted primary translation product of sNMT has a high degree of similarity to NMTs from other species (Fig. 2). At the amino acid level there was 99.5% identity to hNMT-1 and 95% identity to bovine brain NMT partial sequences [47]. The bovine NMT (spleen and cardiac muscle NMT amino acid sequences are identical) showed 47–50% similarity to Saccharomyces cerevisiae, Candida albicans, Histoplasma capsulatum, Cryptococcus neoformans, Caenorhabditis elegans [68] and Drosophila [69] NMTs (Fig. 2). The sNMT and hNMT differ at two amino acid positions, at position 263 (Lys in sNMT, Thr in hNMT) and at position 274 (Ser in sNMT, Thr in hNMT) [45]. Recently, the second distinct mammalian NMT cDNA (hNMT-2) has been cloned from a human liver library, as well as the cloning of the respective mouse homologue for each of the two human NMTs [70]. The mouse and human versions of each NMT are highly homologous, displaying greater than 95% amino acid sequence identity [70]. Comparison of the hNMT-1 and hNMT-2 Table 1. Kinetic properties of NMT isozymes Source of the enzyme

Substrate

Km (µM)

Ref

Bovine spleen

pp60src A kinase MARCKS M2 gene segment

41 250 50 111

53 53 53 53

Bovine cardiac muscle

pp60src M2 gene segment A kinase

16 50 100

62 62 62

Bovine brain

pp60src M2 gene segment A kinase

17 25 64

46 46 46

Rat liver

pp60src A kinase

19 105

40 40

Yeast

pp60src A kinase

40 60

40 40

Rat brain

pp60src A kinase

60 225

63 63

Rat brain

pp60src

20

64

Human

pp60src A kinase

40 120

Rabbit descending colon

pp60src A kinase

40 37

54 54

Rabbit ascending colon

pp60src A kinase

25 89

54 54

65–67 65–67

Fig. 2. Analysis of the predicted primary structures of orthologous NMTs. The multiple sequence alignment was generated using the CLUSTAL PC/GENE program. The conservation is represented by colors: primary (red, 100%), secondary (green, 50%) and tertiary (purple, 25%). The primary structures of C. albicans, S. cerevisiae, C. neoformans, H. capsulatum, hNMT and C. elegans were taken from Zhang et al. [68]. Bovine NMT (bovine spleen and bovine cardiac muscle NMT amino acid sequences are identical, hence we named these NMTs as Bovine was taken from Raju et al. [45]. Human NMT 1 and 2 were taken from Giang and Cravatt [70]. Drosophila NMT was taken from Ntwasa et al. [69].

139

140 proteins revealed reduced levels of sequence identity (76– 77%) suggesting two distinct families of NMT. Human NMT1, bovine NMTs showed 76% identity with hNMT-2 (Fig. 2). The phylogenetic tree analysis of the NMT family reveals that these proteins can be grouped into seven distinct families based on the conservation of amino acids between various NMT species of divergent origin (Table 2). Bovine NMT, hNMT, hNMT-1, M-NMT-1 can be grouped as subfamily-1, hNMT-2 and M-NMT-2 can be grouped as subfamily-2, Drosophila as subfamily-3, C. elegans as subfamily-4, S. cerevisiae and C. albicans as subfamily-5, H. capsulatum as subfamily-6 and C. neoformans as subfamily-7 (Fig. 3). These subfamilies are grouped together in Fig. 2 on the basis of similarities observed on close examination of the sequence alignments. It is also interesting to note that the essential amino acids which are involved in the catalytic activity of NMT are highly conserved in various species of divergent origin [44, 68, 69]. Though the species might have diverged several million years ago, there is a high degree of conservation of NMT structure at the nucleotide as well as at the amino acid level. This high conservation might have some evolutionary and functional implications concerning this gene family. The functionality of the sNMT clone was studied by expressing this cDNA using a histidine-tagged vector which resulted in the production of active NMT [71]. The purified protein has an apparent molecular mass of 50 kDa on SDSPAGE [71] similar to the molecular mass reported for the NMT purified from bovine spleen [53]. The enzyme exhibited a specific activity of 46 nmol/min/mg protein in the presence of pp60src derived peptide substrate, similar to the specific activity reported for the purified enzyme from the bovine spleen [53]. The recombinant sNMT has a lower Km for pp60src (40 µM) than for cAMP-dependent protein kinasederived peptide substrate (200 µM), similar to the values reported for the spleen enzyme purified previously [53].

Fig. 3. Phylogenetic tree of N-myristoyltransferase amino acid sequences. For details see Raju et al. [45].

5′ and 3′ deletions of sNMT To further characterize the essential regions of this enzyme, deletions were generated at the N- and C-terminal ends [45]. Table 3 shows the effect of deletions on sNMT activity. Deletion of 51 amino acids from the N-terminus did not effect the enzyme activity, suggesting that these residues are not critical. However, deletion of only 21 amino acids from the N-terminus caused a 3-fold increase in enzyme activity.

Table 2. Phylogentic tree analysis of NMT family* HNMT-2 H-NMT-2 M-NMT-2 H-NMT-1 M-NMT-1 B-NMT HNMT DRO C-ELE S-CER C-ALB H-CAP C-NEO

M-NMT-2

H-NMT-1

M-NMT-1

B-NMT

HNMT

DRO

C-ELE

S-CER

C-ALB

H-CAP

C-NEO

90

72 68

72 67 97

67 62 83 83

67 62 83 83 99

50 46 50 50 60 60

46 42 44 44 49 50 46

33 31 33 34 39 38 36 35

36 33 34 34 40 40 38 36 52

36 34 35 35 34 34 33 34 41 38

36 34 35 35 40 40 36 36 44 43 37

*The sequences are labeled as in Fig. 2. Bootstrap analysis values as percentage are shown.

141 Presently the functional significance of this activation is not known. However, the results suggest that this could involve deletion of an inhibitory domain. Deletion beyond 52 amino acids resulted in a loss of enzyme activity, whereas all 3′ deletions resulted in the loss of enzyme activity [45]. These studies suggest that sNMT cannot tolerate any deletions at the C-terminal end. A glycine residue near the C-terminal end is highly conserved among the various orthologous NMTs (Fig. 2). C-terminal deletion including Gly412 in sNMT resulted in complete loss of enzyme activity (Table 3). Human and yeast NMTs contain a Gly residue at the C-terminus and Gly to Asp or Lys mutagenesis in these NMTs produced a marked reduction in their activities in E. coli as well as in a temperature sensitive myristic acid auxotroph in S. cerevisiae [51]. Recently the crystal structure of NMT has been described with bound myristoyl-CoA and peptide substrate analogs [72]. A novel catalytic mechanism has been proposed involving deprotonation of the N-terminal ammonium of a peptide substrate by the enzyme’s C-terminal backbone carboxylate [72]. As reported earlier, the yeast enzyme could

withstand considerable deletions and substitutions at the Nterminal end and retain enzyme activity [73]. The highest sequence divergence was observed at the N-terminal region of various orthologous NMTs. Previous deletion studies of yeast NMT revealed that its N-terminal 59 residues do not play a role in catalysis and further suggested that they may function as a targeting signal, allowing the enzyme to access myristoyl-CoA pools produced by activation of exogenous C14:0 [74]. Recently it has been shown that amino-terminal region of NMT is involved in ribosomal targeting signal [75]. This N-terminal sequence divergence is most evident with H. capsulatum NMT which contained an apparent extension of 71 residues compared to other enzymes [74].

Two genes encode bovine NMT Previously a single human NMT cDNA has been isolated and characterized [51], biochemical evidence has indicated the presence of several distinct NMTs in vivo often varying in

Table 3. 5′ and 3′ deletions of sNMT cDNA Deletion C*

S A

S 5′ A

S 5′ A

S 5′ A

S 5′ A

Primer

Expressed protein (between amino acids)

NMT activity (%)

5P CCC AAG CTT ATG AAC TCT TTG CCA 5P CCG GAA TTC GTG GTG GGG GTC CAT TTC

1–416

100

21–416

337

52–416

78.0

134–416

3.6

149–416

3.6

5P CGC GGA TCC ATG GTC AGG GAC CTG CGA A 5P CCG GAA TTC GTG GTG GGG GTC CAT TTC 5P CGC GGA TCC ATA ACA CAD ATG GTC CTG TG 5P CCG GAA TTC GTG GTG GGG GTC CAT TTC 5P CGC GGA TCC ATC GAG TGG TCT CAA GTC GG 5P CCG GAA TTC GTG GTG GGG GTC CAT TTC 5P CGC GGA TCC ATC CGG CAA TCC ATA TC 5P CCG GAA TTC GTG GTG GGG GTC CAT TTC

3′

S A

5P CAG CAG AGA GGA TCC AGG 5P CCG GAA TTC TTA GTT GCC ATC CCC AAT GCC

9–393

3.6

3′

S A

5P CAG CAG AGA GGA TCC AGG 5P CCG GAA TTC TTA CTC CAT GAG ATC

9–377

2.4

S – sense primer; A – antisense primer; C* – results expressed as percent of control NMT. For experimental details see Raju et al. [45].

142 either apparent molecular weight (48–67 kDa) and/or subcellular distribution [42, 47–50, 53] and these earlier studies speculated that NMT activity in vivo is likely derived from a single copy gene [51, 76] . Additional studies were performed to determine whether the bovine genome supported expression of multiple NMTs from more than one genomic sequence. To address this question genomic DNA was digested with BamHI, BglII, and EcoRI and probed with a 1.24 kb sNMT cDNA (Fig. 4). The results indicated that BamHI yielded a prominent hybridization band (12 kb), whereas BglII yielded two equal intensity bands with different sizes (> 15 and 9 kb). Digestion with EcoRI produced three bands, one with higher intensity (14 kb) and two with low intensity (5.3 and 3.8 kb). Of the three restriction enzymes used, only BamHI has one recognition site close to the initiation codon and none for BglII and EcoRI. The different restriction pattern with BglII and EcoRI may result from the presence of these sites within an intron or the different sized restriction fragments may represent separate genomic copies. To address this question of whether the multiple bands represent a single copy gene or two copies of NMT, 5′ and 3′ probes were employed. Genomic Southern analysis with the 5′ probe revealed identical bands similar to the bands observed with full length spleen cDNA probe [45]. The 3′ probe yielded a similar hybridization pattern as with the full length spleen cDNA and 5′ probe except the middle band (5.3 kb) in the EcoRI digest was missing (Fig. 4). These results indicate that the prominent band obtained with BamHI could be two gene fragments of similar size and a restriction length polymorphism (different size fragments) observed with regard to BglII and EcoRI perhaps represent two different gene copies. The missing EcoRI fragment using the 3′ probe suggests that this fragment could arise from the N-terminus of one gene. Double digests of either BamHI and BglII or BamHI and EcoRI indicated the presence of a 12 kb which was similar to the band observed with BamHI digestion alone, whereas the pattern of BglII or EcoRI digestion (in combination with BamHI) were altered compared to either BglII or EcoRI alone [45].

Fig. 4. Southern blot analysis of genomic DNA. Genomic DNA was digested with the indicated restriction enzymes and the reaction products were separated by agarose gel electrophoresis. Blots were probed with sNMT cDNA probe (A) and a DNA fragment from the 3′ region of sNMT cDNA (B). For experimental details see Raju et al. [45].

These studies suggest that there could be internal restriction sites for BglII and EcoRI in one of the BamHI genomic copies. Therefore, these results suggest the possibility for the presence of two homologous gene copies in the bovine nuclear genome [45]. The genomic organization of bovine NMT is currently under investigation. Isolation of cDNAs from various tissues and characterization of these cDNAs would provide a molecular basis for the observed isozymic pattern or mechanism of alternative splicing/post-translational processing. Recently, the isolation of two cDNAs in mouse and human with several single nucleotide changes within the coding sequence clearly suggests that there is more than one NMT gene [70]. In light of the two reported NMT cDNAs, the existence of more than two isoforms can not be explained. Alternative forms of NMT would, therefore, have to be due to either post-translational modifications or post-transcriptional changes such as alternative RNA splicing or trans splicing. There is some recent evidence indicating that alternative splicing is involved in generating different NMT isoforms [77]. Recently we have reported the genomic organization of human NMT-1 [78] which is present on human chromosome 17 (17q21.1). The human NMT-1 gene is composed of 11 exons and 10 introns with consensus AT/GT boundaries [78]. We have also suggested the role of trans-splicing in the generation of NMT isoforms [78]. An interesting observation from this study is that although Bos taurus and Homo sapiens diverged several million years ago, there is a high degree of conservation of NMT structure at the nucleotide as well as at the amino acid level. This conservation could extend into the 5′ and 3′ untranslated sequences and may have some evolutionary and functional implication concerning the gene itself.

Regulators of NMT function The cDNAs encoding yeast [79], human [51, 77], bovine [45, 62], C. elegans [68] and Drosophila [69] NMTs have been reported. However, very little is known with respect to their regulation. Variable tissue distribution of NMT activity has been reported previously [32, 45, 63, 80]. It is not clear whether the observed differences are due to the posttranslational modification (activation/inhibition) of NMT activity or due to the reduced/increased transcription of NMT mRNA. RNA blot hybridization of total cellular RNA prepared from bovine brain, heart, spleen, lung, liver, kidney, and skeletal muscle probed with sNMT cDNA revealed a single reactive band (approximately 1.7 kb) (Fig. 1) suggested a single-sized transcript in all these tissues [45]. Western blot analysis of various bovine tissues (Fig. 1) with hNMT peptide antibody indicated that a prominent

143 immunoreactive band was observed in all tissues with an apparent molecular mass of 48.5–50 kDa. Additional immunoreactive bands were observed in brain (84 and 58 kDa), lung (58 kDa), and skeletal muscle (58 kDa). Further studies were carried out to determine how the protein levels observed in different tissues relate to the enzyme activity. Analysis of tissue extract for NMT activity demonstrated that brain contains highest activity when compared to spleen, lung, kidney, heart, skeletal muscle, pancreas and liver whereas Western analysis of NMT from these tissues showed reactive bands of similar intensity (Fig. 1). It appears, therefore, that mRNA and protein expression in various bovine tissues do not correlate with observed NMT activities, suggesting the presence of regulators of NMT activity [45]. These experiments led us to the discovery of various regulators of NMT function. Identification and characterization of some of the regulators of NMT are discussed in the following section.

N-myristoyltransferase activator protein form bovine brain Careful monitoring of the purification of bovine brain NMT led us to the discovery of an N-myristoyltransferase activator (NAF45) [83]. NAF45 was identified from the fraction of NMT activity which failed to bind to the phosphocellulose column, termed as NMT·PU and represented 32% of the total NMT activity applied to this column [83]. The NAF45 and NMT activities were resolved by mono Q column chromatography from the NMT·PU fraction. This resolution resulted in the loss of NMT·PU activity [83]. These results indicated the existence of an activator of NMT. NAF45 was a nondialyzable, heatlabile molecule with an apparent molecular mass of 45 kDa [83]. The physiological role of NAF45 is at present under investigation in our laboratory.

Bovine cardiac muscle NMT N-myristoyltransferase inhibitory protein from bovine brain An NMT inhibitor was discovered in our laboratory during subcellular fractionation of bovine brain crude extract [81]. Reconstitution of the bovine brain particulate and soluble fractions resulted in inhibition of the elevated cytosolic NMT activity [81]. These results suggested the existence of a putative inhibitor of NMT activity located in the particulate fraction and was proven to be a protein by its susceptibility to the proteases trypsin and chymotrypsin [81]. Proteolytic inactivation requires prior removal of the lipid from the particulate fraction [81]. The NMT inhibitor was purified to near homogeneity by heat treatment, solvent extraction and Sephacryl S-300 gel filtration chromatography. The inhibitor protein had an apparent molecular mass of 92 kDa by gel filtration and 71 kDa by SDS-PAGE, suggesting that the protein is monomeric [81]. Therefore, we designated this protein inhibitor as NIP71 (71 kDa N-myristoyltransferase inhibitor protein). NIP71 does not bind myristoyl-CoA, or interfere with binding of myristoyl-CoA to NMT and possesses no thioesterase, demyristoylase or protease activity [81]. NIP71 potently inhibited highly purified NMT II activity (IC50 23.7 nM), whereas other forms of bovine brain NMT were insensitive to NIP71 inhibition [46]. NIP71 also inhibited bovine spleen [53], human [67], and rat colonic tumor [82] NMTs. Inhibition by NIP71 at subsaturating concentration of peptide (20, 35 and 50 µM) and myristoyl-CoA (0.27 µM) indicated that the NIP71 inhibited NMT activity in a sigmoidal manner [46]. Therefore, a small change in the concentration of NIP71 would result in a large change in the NMT activity. This study suggests that bovine brain may possess a potent and delicate on/off switch to control NMT activity.

Several laboratories have reported the existence of low NMT activity in crude homogenates of rat cardiac muscle extracts [32, 45, 64, 78]. We have also reported recently low NMT activity in the cytosolic fractions of bovine cardiac muscle [62, 84]. The low NMT activity in bovine cardiac muscle could be due to the low expression of the NMT gene or could be due to the presence of regulators of enzyme activity. In order to address this issue, gene expression was carried out employing sNMT cDNA as a probe [62]. The results indicated a 1.7 kb hybridizing band suggesting the NMT gene expression in cardiac muscle (Fig. 1). This observation is further confirmed by RT-PCR on bovine cardiac muscle poly A (+) RNA employing sNMT cDNA specific primers resulted in the amplification of an expected 1.24 kb PCR product [62]. The single long open reading frame of bovine cardiac muscle NMT cDNA specifies a protein of 416 amino acids with a predicted mass of 46,686 Da [62]. An alignment of the cNMT sequence with those of NMT from other species reveals a high degree of conservation throughout the primary structures (Fig. 2). The functionality of the cDNA clone was studied by expressing this cDNA, which resulted in the production of an active 50 kDa NMT [62]. Immunohistochemical localization of NMT from autopsy specimens of human heart revealed a strong staining of myocardial muscle fibres with polyclonal antibody raised against hNMT [62]. The immunoreactivity was observed throughout the cytoplasm of the myocytes [62]. Ultrastructural localization of NMT demonstrated cytoplasmic with occasional mitochondrial and myofilament (Fig. 5) distribution suggesting an important role of this protein in cardiac muscle.

144 Myristoyl-CoA binding protein from bovine cardiac muscle We have identified and characterized a myristoyl-CoA binding protein (MCBP) from bovine cardiac muscle which could regulate the myristoylation reaction in cardiac muscle [84]. Absence of correlation between the activity studies and gene expression (mRNA and protein) could be due to the presence of MCBP in the cardiac muscle. Myristoyl-CoA binding protein was purified using various column chromatographies: hydroxylapatite (HTP), DEAE Sepharose CL6B and Sephacryl S-300 gel filtration [84]. The purified protein exhibits an apparent molecular mass of 50 kDa on SDS-PAGE. Exogenous peptide substrate myristoylation was observed with the cytosolic fraction and to some extent, with the first HTP column fraction (Table 4). Purified protein was incubated with radiolabeled myristoyl-CoA shows the linear, time-dependent incorporation of radiolabeled myristoyl-CoA into the protein, suggesting the formation of acyl-protein complex. Substitution of [3H]palmitoyl-CoA as the acyl donor resulted in less than 7% of the radioactivity being incorporated into the protein as compared to myristoyl-CoA. The CoA itself did not bind to the protein at a similar concentrations to that of myristoyl-CoA, however, a 2500fold higher concentration of CoA is required for the complete inhibition of myristoyl-CoA binding to the protein. However, myristic acid did not serve as acyl donor for the protein [82].

Evidence for high affinity (covalent) acyl-protein complex

Table 4. Comparison of NMT activity in various fractions obtained during purification Fraction

Protein mg/ml

Cytosol 15.0 HTP column 8.3 DEAE Sepharose CL-6B 3.0 (unbound) Sephacryl S-300 gel 0.11 filtration HTP column 0.15

Activity pmol/min/ml

Specific activity U/mg

0.0 (33.0) 107.0 (142.0) 110.0 (110.0)

0.0 (2.2) 13.0 (17.0) 36.0 (36.0)

307.0 (265.0)

2842.0 (2453.0)

464.0 (461.0)

3093.0 (3073.0)

Myristoyltransferase activities was measured in each fraction in the presence and absence of cAMP-dependent protein-kinase peptide. Parentheses indicate the enzyme activity in the presence of peptide. For experimental details see Raju and Sharma [84].

define the chemical nature of the interaction between the MCBP and its acyl-ligand, an identical series of samples (purified radiolabeled MCBP) were subjected to electrophoresis through denaturing SDS-polyacrylamide gels. These gels were subsequently stained with Coomassie Brilliant Blue, destained in acetic acid/isopropyl alcohol and washed in 1 M hydroxylamine at pH 7.0 [84]. The gels were treated with EN3HANCE (Dupont, NEN), dried and subjected to fluorography. The association of [1-14C]myristoyl-CoA with protein was sensitive to hydroxylamine pH 7.0. This result is consistent with an acyl-protein linkage occurring via an ester bond, since the radioactivity was completely abolished in the presence of hydroxylmaine [84]. These results exclude the possibility of an endogenous/automyristoylation, since myristoylation occurs via an amide bond [12, 19, 24, 25].

Thioester and reactive oxyester bonds are susceptible to hydroxylamine-mediated hydrolysis at pH 7.0. To further Effect of cytosol on protein-acyl-CoA complex in MCBP

Fig. 5. Immunolocalization and electronmicroscopy. Ultrastructural localization of NMT in human cardiac muscle by using immunogold labelling technique demonstrates (A) cytoplasmic (arrow), myofilament (circle) and (B) mitochondrial (arrowhead) immunoreactivity (EM magnification × 50,000). For experimental details see Raju et al. [62].

As presented in Table 4 indicated that the cytosolic fraction was capable of transferring the myristate moiety from myristoyl-CoA to exogenous peptide substrates. The nature of the cytosolic fraction was further investigated in detail. Purified MCBP was incubated with [1-14C]myrisotyl-CoA in separate tubes and aliquots were removed at selected times. At 10 min, cytosolic fraction was added to one tube and buffer to other. Aliquots were removed at selected times up to 120 min. The vial which did not contain added cytosol exhibited the formation of a protein-myrisotyl-CoA complex, while the tube which contained added cytosol exhibited deacylation (Fig. 6). These results suggest that the cytosolic fraction may contain thioesterases/proteinases which could modulate the acylation reaction in vivo [84]. Very recently it has been reported that porcine phospholipase A2 contained thioesterase and deacylase activities [85]. The physiological consequence of acylation and deacylation of proteins have been shown to be involved in

145 regulating the activity and cellular localization of many proteins [85]. The deacylation process has been involved in the fusion between viral and cellular membranes [86] and in the release of virus particles from infected cells [87], showing a decrease in both activities when proteins are deacylated by hydroxylamine [88]. Our studies indicated that the myristoylation reaction in bovine cardiac muscle could be regulated by the relative distribution of NMT, myristyl-CoA, MCBP and acyl-synthetases. The regulation of acyl-protein complexes in signal transduction system has been hampered due to lack of knowledge about the specific deacylases. In the near future purification and characterization of deacylases will greatly facilitate our understanding of the regulation of protein myrisotylation, especially the regulation of acylprotein complexes.

Regulation of NMT activity by Ca2+ dependent cysteine protease m-calpain Many short lived proteins which are devoid of proteolytic activity contain PEST sequences which are segments along the polypeptide chain that are rich in proline (P), glutamic acid (E), serine (S) and threonine (T) [89–91]. These designated PEST sequences are believed to be putative intramolecular signals for rapid proteolytic degradation [90, 91]. PEST regions are considered to be recognized by specific proteases, particularly calpains [90, 91]. PEST sequences vary in length from 12–100 amino acids and can reside anywhere along the molecule [90, 91]. The strength of the PEST sequence is determined by the PEST score which range from –45 to +50. A PEST score of < 0 but > –5 represents a weak PEST region, but a value > +5 indicates a very strong

Fig. 6. Deacylation of protein-myristoyl-CoA complex. Purified bovine cardiac muscle myristoyl-CoA binding protein (¡-¡) and cytosolic fraction (¨-¨) was incubated with [1-14C]myristoyl-CoA in the absence of peptide substrates. Aliquots were removed at various time intervals. Arrow indicates the time of addition of cytosol and progress of the reaction was followed (G-G). For experimental conditions see Raju and Sharma [84].

PEST motif [90, 91]. The PEST regions contain concentrations of negatively charged residues surrounded by clusters of basic amino acids and are most likely located on the surface of proteins [92, 93]. Known PEST proteins include oncogene products (c-Myc, c-Fos and c-Myb), enzymes (hydroxymethyl glutaryl-CoA reductase, ornithine decarboxylase), p53, nuclear factor κB and components of signal pathways, calmodulin binding proteins, kinases and various receptors [89, 91]. Recently we have also demonstrated the generation of active calmodulin-independent phosphodiesterase from brain calmodulin-dependent phosphodiesterase by m-calpain [94]. By computer analysis (http:// www.icnet.uk/cgi-bin/ runpest.pl) of the amino acid sequence, eight PEST regions (P1: 14–27, P2: 35–47, P3: 66–86, P4: 109–122, P5: 192– 209, P6: 281–293, P7: 293–330, P8: 346–363) were found in the cardiac muscle NMT (cNMT) [62]. The PEST scores of P1 to P8 were –27.0, –12.2, –5.7, –22.6, –29.4, –4.7, –7.4 and –17.3, respectively. To examine the effects of NMT processing by m-calpain, cNMT was pre-incubated with calpain either in the presence of Ca2+ or EGTA or calpastatin. NMT activity was abolished by m-calpain in a time dependent manner in the presence of Ca2+ (Fig. 7) [62]. Western blot analysis indicated that NMT protein was degraded by calpain in a time dependent manner (Fig. 7). Abolishment of NMT activity (Fig. 7) and degradation of NMT protein (Fig. 7) by m-calpain was efficiently inhibited by calpastatin, an inhibitor of calpains [62, 95]. Both calpastatin and calpains are ubiquitously expressed in different tissues and are thought to play an important role in Ca 2+-dependent regulatory systems [92, 93]. Calpains cause limited proteolysis of substrates resulting in the modification of activity rather than inactivation or complete degradation of substrates [92, 93]. However, in our study, cNMT protein expression was completely degraded by calpains which resulted in the loss of enzyme activity [62]. The complete proteolytic degradation of cNMT could be due to the presence of random PEST sequences. It is interesting to note that the primary translational product of cNMT has a high degree of similarity (> 95%) to spleen and human NMT and 47–66% identity to yeast, Drosophila and C. elegans NMTs (Fig. 2). These observations suggest that PEST sequences are also present on these NMTs from diverse organisms. Degradation of NMT protein and modulation of NMT activity by calpains may offer an in vivo downregulatory mechanism. Currently, there are contradictory arguments with regard to the intracellular localization of calpains [90]. There have been a number of papers demonstrating that calpains are located mainly in the cytoplasm, and very scarcely, in the nucleus [96– 98]. Previously, we reported on the cytosolic localization of NMT in cardiac muscle [99]. Localization of calpain and calpastatin in the cytoplasm may have regulatory role on the turnover of NMT.

146 Breakdown/degradation of NMT by calpains may offer a potential therapeutic target. The involvement of NMTs in apoptosis is yet to be determined. Recently, it has been shown that the NMT substrates lck, src and abl inhibit apoptosis in a concentration dependent manner [113]. The placement of myristic acid on the amino-terminal sequence of these proteins allows it into the cell membrane where it participates in signal transduction. Without myristoylation these proteins do not function during signal transduction [12, 19, 24, 25]. Therefore, there is evidence to support the hypothesis that myristoylation of proteins may be involved in apoptosis. µ- and m-calpain have been reported to be implicated in myocardial stunning and other cardiac injuries [114]. It has been reported that during ischemia and reperfusion elevated Ca2+ flux results in enhanced calpain activity leading to the proteolysis of several important proteins [115]. During hypoxia increased calpain activity leads to irreversible cell membrane degradation resulting in irreversible cellular injury [92]. Availability of cDNA clones of cNMT [62] would greatly facilitate studies of the in vivo regulation of NMT by calpains and calpastatin as well as studying the role of NMT in apoptosis/cardiac functions.

Effects of L-histidine and its structural analogs on hNMT activity

Fig. 7. (A) Effect of m-calpain on the activity and protein degradation of NMT. Recombinant NMT was incubated with partially purified m-calpain either in the presence of Ca2+ or EGTA at indicated times. NMT activity was carried out in the presence of cAMP-dependent protein kinase peptide substrate. At the various time intervals indicated, an aliquot was removed for SDS-PAGE and immunoblot (inset); (B) Effect of calpastatin on NMT. NMT was incubated with partially purified m-calpain either in the presence of calpastatin or Ca2+ or EGTA at indicated times. NMT activities and immunoblotting (inset) was carried out. For experimental conditions see Raju et al. [62].

Recently, it has been shown that alpha-spectrin (fodrin) is broken down by calpain and ICE like protease in apoptotic cells, suggesting that both protease families participate in the expression of neuronal apoptosis [100]. It is tempting to speculate that calpain and calpastatin are mimicking the players of apoptosis in a way similar to Ced-3 (C. elegans) and ICE (vertebrates), the cleavage proteases and Ced-9 (C. elegans) and Bcl-2 (humans), inhibitors of apoptosis [101– 103]. NMT is a potential target for anti-viral [104–106], antifungal [107–109], and anti-neoplastic [82, 110–112] therapy.

NMT activity was found to be activated by L-histidine in a concentration-dependent manner (Fig. 8). Similar results were obtained with D-histidine [116] suggesting that the NMT activation by either D-or L-histidine is due to the imidazole moiety. Previously, we have demonstrated the activation of sNMT by imidazole buffer [53]. However, both D-and L-histidine are much more effective activators of NMT than imidazole, suggesting that both the imidazole and either α-amino or α-carboxyl moieties were needed for activation. That histidine-O-methylester inhibited NMT activity implicated the α-carboxyl group of histidine [116]. The fact that several other amino acids (L-tryptophan, L-isoleucine, and glycine) failed to influence NMT activity suggested that ‘full’ NMT activation depended upon both an imidazole moiety and a nearby carboxyl-residue [116]. This notion is supported by the finding that NMT activity was also inhibited in a concentration-dependent manner, by L-histidinol and histamine with a half-maximal inhibition of 18 and 1.5 mM, respectively (Fig. 8). Kinetic analysis was carried out in the absence and presence of L-histidine, L-histidinol and histamine using peptide derived from the N-terminal sequence of cAMP-dependent protein kinase derived peptide substrate [116]. These results indicate that L-histidinol and histamine appear to be noncompetitive inhibitors, because inhibition of the NMT activity by these compounds does not change the Michaelis constant, whereas Vmax is changed. The

147 inhibition of NMT by L-histidinol can be reversed by an excess L-histidine, suggesting that NMT is not covalently modified by the histidine analogs [116]. Similar results were obtained with histamine [116]. The results further suggest that L-histidine, L-histidinol and histamine could be competing for the common site on NMT. Recently, we have reported the possible inhibition of NMT activity by L-histidinol and histamine by competing His-293 in NMT [116] were based on the previously proposed hNMT reaction mechanism [75]. Weston et al. [117] have shown that His-302 (C. albicans) is located remote from the active site on the opposite side of the molecule. Recently the structure of NMT has been described with bound myristoyl-CoA and peptide substrate analogs [72]. In this structure, the main chain amide groups of Phe 170 and Lue 171 constitute the oxyanion hole [72]. As a consequence, the thioester carbonyl of myristoyl-CoA becomes polarized with a partial positive charge on the carbon setting the stage for nucleophilic attack [72]. Previously it has been shown that imidazole-substituted dipeptide amides are selective inhibitors of C. albicans NMT and that these agents have antifungal activity [109]. Lhistidinol has been shown to modulate cancer drug toxicity [118], both in vitro (optimum concentration 1 mM in

Fig. 8. Effect of L-histidine, L-histidinol and histamine on hNMT. Human NMT was incubated with either L-histidine (0–40 mM , ) or L-histidinol (0–34 mM, ) or histamine (0–4 mM, ) in the presence of cAMP-dependent protein kinase substrate. For experimental details see Raju et al. [116].

conventional tissue culture media) and in DBA/2J mice bearing either L1210 leukemia [119, 120], P815 mastocytoma [121], or P388 leukemia [122], as well as in C57/BL mice bearing disseminated B16f10 melanoma [123]. L-Histidinol has also been shown to reverse several types of drug resistance [124], including a unique form of resistance observed in MDCK-T1, a tumorigenic epithelial cell line [124]. The data presented here suggest that various imidazoles may elicit effects upon myristoylation as well. It is also of interest that the drug modulating action of L-histidinol on tumor cells is eliminated by excess L-histidine, both in tissue culture, and in tumor bearing animals [116, 123]. Myristoylation has been identified as a potential target for the development of chemotherapeutic agents [111]. The results presented here suggest that histidine analogs and other imidazole derivatives may be useful as potential inhibitors of protein myristoylation.

Elevated NMT activity in colorectal adenocarcinomas Colorectal cancer is one of the leading causes of cancer death in North America. Since treatment of colonic cancer remains difficult because of the lack of effective chemotherapeutic agents, it is important to continue to search for cellular functions that can be disrupted by chemotherapeutic drugs in order to inhibit the development or progression of this disease. Modification of proteins by myristoylation has been recognized as important in the function of various, viral, oncogenic, and signal transduction proteins and thus has been proposed as a target for chemotherapeutic drug design [104– 112]. The importance of the myristoylation of proteins in tumorigenesis was first suggested by studies demonstrating that myristoylation of the viral oncogene pp60v-src is required for membrane association and cell transformation [125–127]. Myristoylation is also required for association of pp60v-src peptide with the 32 kDa membrane receptor [22]. Blockage of the myristoylation of pp60 c-src in colonic cell lines depressed colony formation, cell proliferation and localization of pp60 c-src to the plasma membrane [128]. The function and subcellular localization of other proteins involved in signal transduction are variously altered by mutations resulting in loss of myristoylation [129–130]. Therefore, there is evidence to support the hypothesis that myristoylation of proteins may be involved in the pathogenesis of cancer. The activity of NMT in cancer tissue compared with normal tissue has not been investigated previously. We were the first to demonstrate the elevated levels of NMT activity in colonic tumors [82]. In this study we have used the azoxymethane-induced rat model of colon carcinogenesis to investigate the activity of NMT in colonic tumors. This model has been used extensively to characterize changes in various enzymes during colon carcinogenesis [131].

148 Myristoyltransferase activity in the rat colonic tumors was higher than NMT activity in normal and normal-appearing colonic mucosa. NMT activity in colonic mucosa in control rats (those not given injections of azoxymethane) was similar to the NMT activity in normal appearing mucosa in the tumor bearing rats (Table 5). All rat colonic tumors had higher NMT activity than the corresponding normal-appearing mucosa. Overall, NMT activity tended to decrease with the distance of the tumor from the rectum (r = –0.544; p = 0.001). There was no linear association between the size of the tumors and NMT activity [82]. Increased NMT activity was also observed in human colonic tumors and was predominantly cytosolic [82]. Increased NMT activity was not observed in Crohn’s disease or volvulus, suggesting that elevated NMT activity is not a nonspecific response occurring in an inflammatory conditions or noncancerous lesions [82]. NMT activity was markedly elevated in rat adenomatous polyps and stage B1 tumors (Table 6). This observation suggest that NMT activity is elevated in the early stages of colonic carcinogenesis, a finding that may be of diagnostic and/or prognostic value. Adenomatous polyps consist primarily of epithelial cells, whereas more advanced tumors contain many fibrous and inflammatory cells. Therefore NMT activity may differ in various cell types. We have observed higher NMT activity in rat colonic tumors closest to the rectum. In human colonic cancer, proximity to rectum has been associated with poor prognosis in several studies [132–134] but not in all studies [135]. Although it is tempting to speculate that high NMT activity in tumors may be related to poor prognosis, it is not known whether the relationship of tumor location and prognosis holds true in the rat model. Because colonic cancer is thought to originate in the epithelial cells [136], we assayed NMT activity in the epithelial mucosa separated from the underlying smooth muscle layers. Our finding that NMT activity is elevated in colonic cancer suggesting that the activity or expression of this enzyme may be regulated in normal colon. The possibility of regulation of myristoylation of proteins has been suggested by previous studies [36, 137,

Table 5. N-myristoyltransferase activity in colonic mucosal homogenates and azoxymethane-induced rat colonic tumors

Tissue

No

N-myristoyltransferase activity* U/mg protein U/gm tissue

Normal mucosa from control rats Normal-appearing mucosa from injected rats Colonic tumors

3

0.34 ± 0.02

7.76 ± 0.89

7

0.43 ± 0.08

8.86 ± 1.86

35

1.69 ± 0.15†

47.99 ± 5.06†

*Mean ± S.E. N-myristoyltransferase activity was determined using cAMPdependent protein kinase-derived peptide as substrate; †Significant difference from normal and normal-appearing mucosa (p < 0.05). For experimental details see Magnuson et al. [82].

Table 6. N-myristoyl transferase activity in azoxymethane-induced rat colonic tumors

Stage of tumor

No

N-myristoyltransferase activity* U/mg protein U/g tissue

Adenomatous polyp

5

2.32 ± 0.38

60.43 ± 12.5

Stage A

8

1.39 ± 0.20

31.19 ± 5.67

Stage B1

10

2.11 ± 0.32

65.39 ± 12.59

Stage B2

10

1.35 ± 0.02

40.80 ± 6.52

Stage C2

2

0.90 ± 0.43

33.03 ± 15.09

*Mean ± S.E. N-myristoyltransferase activity was determined using cAMPdependent protein kinase-derived peptide as substrate . For experimental details see Magnuson et al. [82].

138]. These studies demonstrated the modulation of myristoylation of proteins in macrophages by various stimuli [137] and in mononuclear phagocytes by interferon-γ and tumor necrosis factor-α [138]. Demyristoylation of MARCKS protein in brain synaptosomes has been reported [36]. We speculated that elevated NMT activity may be required during colonic carcinogenesis because of increased expression of various proteins that require myristoylation. The levels of myristoylated tyrosine kinases, pp60c-src and pp62cyes , are several-fold higher in colonic preneoplastic lesions and neoplasmas compared with normal colon [139–141]. It would appear that the high level of pp60c-src activity in colon cancer correlates with both an increase in enzyme synthesis and an association of the activated kinase with the cytoskeleton. Since a relationship has previously been established among pp60 src activity and cellular transformation, and an Nmyristoylation-dependent association of both cellular and viral pp60src with cytoskeleton [126, 142], it is possible that the increased synthesis of pp60c-src in colon cancer requires increased levels of N-myristoylation in order to facilitate the N-myristoyl-dependent targeting of newly synthesized pp60c-src to the cytoskeleton. Myrisotylation of pp60c-src has been reported to be required for dephosphorylation of Tyr527 and activation of src kinase activity during mitosis of NIH 3T3 cells [23]. There may also be other myristoylated proteins involved in colon cancer that are yet to be identified. It is also possible, that in the presence of high NMT activity in neoplasia, proteins that are not normally myristoylated become substrates for the enzyme. For example, the myristoylation of normal cellular p21 ras resulted in potent transformation activity [143]. The myristoylation of H-ras and K-ras altered subcellular localization of MAP kinase [144]. High levels of NMT activity might result in aberrant myristoylation of proteins, resulting in modifications to signal transduction pathways.

149 NMT overexpression in human colorectal adenocarcinomas Our previous studies on colon cancer showed elevated levels of NMT activity in rat and human colonic tumors when compared to normal or normal-appearing mucosa [82]. However, it is not yet clear whether there is an absolute increase in the production of NMT or whether the increased activity is due to a confirmational change in the preexisting enzyme. This is a very important question because the overexpression of NMT in colorectal cancer has implications with regard to the development of chemotherapeutic agents. With the advent of NMT antibody and anti-peptide antibodies directed against human NMT, we have addressed this issue. In both normal mucosa and colorectal adenocarcinomas, NMT with a molecular mass of 48.5 kDa was identified with anti-human NMT and anti-peptide antibody [110]. However, the expression of NMT was found to be higher in the colorectal tumors (Fig. 9). This finding was further confirmed by the immunohistochemical studies which showed stronger cytoplasmic staining in the tumors (Fig. 10) [110]. These findings represent the first description of NMT overexpression in colorectal adenocarcinomas. This has implications with regard to (i) design of chemotherapeutic drugs and (ii) prognosis, for instance, in monitoring colorectal cancer recurrence or metastases. These immunological studies also indicated that antibodies to human NMT have potential utility as a diagnostic marker for the detection of colorectal cancer [145].

Inhibition of N-myristoylation of Src oncoprotein by human serum albumin Myristoylation has been identified as a potential target for the development of chemotherapeutic agents. As indicated, inhibitors of NMT could possibly form the basis of a new approach to cancer treatment. In an investigation to find inhibitors of NMT, we discovered, for the first time, that human serum albumin is a potent inhibitor of N-myristoylation of the oncoprotein-derived protein peptide substrate pp60src in vitro [146] by hNMT. Background information indicated [147] that mutant albumin-deficient rats have hyperlipidemia and are susceptible to the induction by various carcinogens of tumors of the bladder, kidney, stomach, intestine, and subcutaneous tissues [148]. Albumin-deficient rats (analbuminemic rats) are mutants resulting from a defect in splicing of the albumin messenger RNA precursor resulting from the deletion of seven bases in the intron HI, which is responsible for the absence of albumin production in these animals [147]. Hyperlipidemia associated with the lack of albumin may change the membrane lipid composition, altering the permeability to certain carcinogens [148]. We have also

Fig. 9. Immunoblot of normal mucosa and colorectal tumor tissue extracts with anti-peptide NMT antibodies. For experimental details see Raju et al. [110].

reported earlier that NMT activity is low in liver cytosol [149]. It is tempting to suggest that low levels of NMT activity could be related to increased levels of albumin in the liver [150]. Albumin content has been reported to be low in the elderly [151], furthermore an increased expression of pp60src in gastric mucosa of aged rats has also been reported [152]. Taken together, these results suggest that albumin may play a key role in reducing oncoprotein transformation and the development of cancer.

Elevated NMT activity in diabetic rat Diabetes mellitus is divided into two major categories; insulin-dependent diabetes mellitus (IDDM) or juvenile onset diabetes, and the more common non-insulin-dependent diabetes mellitus (NIDDM). Perturbations caused by diabetes will frequently impair signal transduction pathways and thus a cell’s ability to respond to a given stimulus. In many cases, sodium orthovanadate (an insulin-like agent) has been observed to alleviate the effect of diabetes both in vivo and in vitro [153–155].

Fig. 10. Immunohistochemistry of normal and tumor tissue. Section with transitional mucosa (on the left) showing a mild to moderate degree of diffuse reactivity compared to the strong tumor reactivity on the right (immunoperoxidase; original magnification, × 120). For experimental details see Raju et al. [110].

150 We have reported that streptozotocin (STZ) induced diabetes (IDDM ) resulted in a j 2.0 fold increase in the liver NMT activity in all subcellular fractions as compared with control animals [151]. In contrast with IDDM, in obese or NIDDM there was a j 4.7-fold lower particulate NMT activity as compared with the control lean rat liver [156, 157]. However, it is interesting to note that this phenomena was observed only in the particulate fraction. The loss of NMT activity in particulate was not due to an increased activity of NMT in the soluble fraction suggesting that there was no simple translocation of active NMT from the membrane fraction to the soluble fraction. Administration of sodium orthovanadate to the diabetic rats normalized liver NMT activity. Further, the particulate rat liver NMT appeared to be inversely proportional to the level of plasma insulin, implicating insulin in the control of myristoylation [156]. The mechanism of action of insulin and the effects of diabetes on NMT are at present unknown.

Protein myristoylation in plants Knowledge regarding the protein acylation in plants is extremely limited. Yet interaction between lipids and protein complexes are known to occur in thylakoid. It has been reported that polar lipids and their fatty acid constituents play a role in the architecture [158], and probably function [159] of the photosynthetic membrane. Spirodela oligorrhiza, an aquatic angiosperm has been reported to be labeled with [3H]myristic acid [160]. Studies on partial purification of wheat germ NMT have been reported to direct myristoylation reaction in cell free translation system [48]. The Pto kinase from tomato [161], as well as several other plant protein kinases have been observed to have putative myristoylation sites, including the closely related Fen kinase from tomato, the ATN1 and APK1 kinases from Arabdopsis thaliana, and several calcium-dependent kinases from maize [162–165]. A myristoylation motif is also present in the derived aminoacid sequence from several cDNAs encoding fatty acyl-CoA synthetase from Brassica napus and from a nodule-specific gene family from Alnus glutinosa [166, 167]. It is interesting to note that several plant signal transduction proteins have putative myristoylation sites, the enzyme which does this modification has not yet been characterized. Antibodies directed against hNMT showed a cross reactivity with tobacco leaf protein extracts. Additionally, NMT assays with crude leaf cell extracts indicated that the biochemical properties of the plant enzyme were comparable to hNMT. These findings allowed us to partially purify tobacco NMT from leaf extracts. Subcellular distribution of NMT activity indicated that NMT activity is present in a major cytosolic fraction (85%) and a minor activity in particulate fraction (3%). Partial purification of tobacco NMT

was achieved through SP-Sepharose fast flow column and a Superose 12 FPLC gel filtration column chromatographies. The specific activity of the partially purified tobacco NMT was found to be 114 pmol/min/mg protein in the presence of cAMP-dependent protein kinase derived peptide substrate. Under native conditions, the enzyme exhibited an apparent molecular mass of 66 kDa, whereas under denaturation conditions the enzyme represented an apparent molecular mass of 50 kDa suggesting a monomeric protein as demonstrated with the advent of a recombinant hNMT antibodies (Raju, Datla, Sharma, unpublished data). Kinetic properties indicated that tobacco NMT has an apparent low Km for pp60src (27 µM) as compared with cAMP-dependent protein kinase derived peptide substrate (67 µM). Results of our study indicate that tobacco NMT has similar properties to mammalian NMT with respect to kinetic, molecular weight and immunological properties. Further characterization of plant NMT in terms of molecular characterization and identification of cellular targets would give a better understanding of the role of protein myristoylation in plant physiology and pathology.

Acknowledgements This work is supported by the Medical Research Council of Canada, the Health Services Utilization and Research Commission of Saskatchewan, and the Heart and Stroke Foundation of Saskatchewan. We thank Ms. Sherry BarJurado, Department of Microbiology, University of Saskatchewan for helping in the multiple alignments for the primary translation products of orthologous NMTs.

References 1. 2.

3.

4.

5.

6.

7.

Cross CA: Glycolipid anchoring of plasma membrane proteins. Annu Rev Cell Biol 6: 1–39, 1990 Dunphy WG, Fries E, Urbani LJ, Rothman JE: Early and late functions associated with the golgi apparatus reside in distinct compartments. Proc Natl Acad Sci USA 78: 7453–7457, 1981 Quinn P, Griffiths G, Warren G: Dissection of the Golgi complex II. Density separation of specific golgi function in virally infected cells treated with monesin. J Cell Biol 96: 851–856, 1983 Schmidt MFG, Schlesinger MJ: Relation of fatty acid attachment to the translation and maturation of vesicular stomatitis and sindbis virus membrane glycoprotein. J Biol Chem 255: 3334–3339, 1980 Berger M, Schmidt MFG: Cell free fatty acid acylation of semliki forest viral polypeptides with microsomal membranes from eukaryotic cells. J Biol Chem 259: 7245–7252, 1984 Slomiany A, Liau YH, Takagi A, Laszewicz W, Slomiany BL: Characterization of mucus glycoprotein fatty acyltransferase from gastric mucosa. J Biol Chem 259: 13304–13308, 1984 Carr SA, Biemann K, Shozo S, Pamelee DC, Titani K: n-tetradeconyl is the NH2-terminal blocking group of the catalytic subunit of cyclic AMP dependent protein kinase from bovine cardiac muscle. Proc Natl Acad Sci USA 79: 6128–6131, 1982

151 8 . Aitken A, Cohen P, Suntikarn S, Williams DH, Culder AG, Smith A, Clee B: Identification of the NH2 terminal blocking group of calcineurin B as myristic acid. FEBS Lett 150: 314–318, 1982 9. Ozols J, Carr SA, Strittmatter P: Identification of the NH2 terminal blocking group of NADH-cytochrome b5 reductase as myristic acid and the complete amino acid sequence of the membrane-binding domain J Biol Chem 259: 13349–13354, 1984 10. Buss JE, Mumby SM, Casey PJ, Gilman AG, Sefton BM: Myristoylated α subunits of guanine nucleotide-binding regulatory proteins. Proc Natl Acad Sci USA 84: 7493–7497, 1987 11. Schultz AM, Henderson LE, Oroszlan S, Garber E, Hanafusa H: Amino terminal myristoylation of protein kinase p60src a retroviral transforming protein. Science 227: 427–429, 1985 12. Boutin JA: Myristoylation. Cell Signal 9: 15–35, 1997 13. Muszbek L, Laposata M: Myristoylation of proteins in platelets occurs predominantly through thioester linkages J Biol Chem 268: 8251– 8255, 1993 14. Marchildon GA, Casnellie JE, Walsh KA, Krebs EG: Covalently bound myristate in a lymphoma tyrosine protein kinase. Proc Natl Acad Sci USA 81: 7679–7682, 1984 15. Semba K, Nishizawa M, Miyajima N, Yoshida MC, Sukegawa J, Yamanashi Y et al.: yes-related protooncogene, syn, belongs to the protein tyrosine kinase family. Proc Natl Acad Sci USA 83: 5459– 5463, 1986 16. Sukegawa J, Semba K, Yamanashi Y, Nishizawa M, Miyajima N, Yamamoto T et al.: Characterization of cDNA clones for the human cyes gene. Mol Cell Biol 7: 41–47, 1987 17. Arbuzova A, Murray D McLaughlin S: MARCKS, membranes, and calmodulin: Kinetics of their interactions. Biochem Biophys Acta 1376: 369–379, 1998 18. Schultz AM, Tsai SC, Kung HF, Oroszlan S, Moss J, Vaughan M: Hydroxylamine-stable covalent linkage of myristic acid in Gïα, a guanine nucleotide-binding protein of bovine brain. Biochem Biophys Res Commun 146: 1234–1239, 1987 19. Johnson DR, Bhatnagar RS, Knoll LJ, Gordon JI: Genetic and biochemical studies of protein N-myristoylation. Ann Rev Biochem 63: 869–914, 1995 20. Liauw S, Steinberg RA: Dephsophorylation of catalytic subunit of cAMP-dependent protein kinase at Thr-197 by a cellular protein phosphatase and by purified protein phosphatase-2A. J Biol Chem 271: 258–263, 1996 21. Resh MD: Specific and saturable binding of pp60 v-src to plasma membranes: Evidence for a myristoyl-src receptor. Cell 58: 281–288, 1989 22. Resh MD, Ling H-P: Identification of a 32K plasma membrane protein that binds to the myristoylated amino-terminal sequence of p60src. Nature 346: 84–86, 1990 23. Bagrodia S, Taylor SJ, Shalloway D: Myristoylation is required for Tyr-527 dephosphorylation and activation of pp60c-src in mitosis. Mol Cell Biol 13: 1464–1470, 1993 24. Raju RVS, Magnuson BA, Sharma RK: Mammalian myristoylCoA:protein N-myristoyltransferase. Mol Cell Biochem 150: 191–202, 1995 25. Bhatnagar, RS, Gordon, JI: Understanding covalent modification of proteins by lipids: Where cell biology and biophysics mingle. Trends Cell Biol 7: 14–20, 1998 26. Kamps MP, Buss JE, Sefton BM: Mutation of NH2-terminal glycine of p60src prevents both myristoylation and morphological transformation. Proc Natl Acad Sci USA 82: 4625–4628, 1985 27. Buss JE, Kamps MP, Gould K, Sefton BM: The absence of myristic acid decreases membrane binding of p60src but does not affect tyrosine protein kinase activity. J Virol 58: 468–474, 1986 28. Cross FR, Garber EA, Pellman D, Hanafusa H: A short sequence in the p60 src N-terminus is required for p60 src myristoylation and

29.

30.

31. 32. 33.

34.

35. 36.

37.

38.

39.

40.

41.

42.

43.

44.

45.

46. 47.

membrane association and for cell transformation. Mol Cell Biol 4: 1834–1842, 1984 Kamps MP, Buss JE, Sefton BM: Rous sarcoma virus transforming protein lacking myristic acid phosphorylates known polypeptides substrates without inducing transformation. Cell 45: 105–112, 1986 Rein A, McClure MR, Rice NR, Luftig RB, Schultz AM: Myristoylation site in pr65gag is essential for virus particle formation by Moloney murine leukaemia virus. Proc Natl Acad Sci USA 83: 7246–7250, 1986 Stuck JW, Lehmann LH, Siegel EG: Acylation of proteins by myristic acid in isolated mitochondria. J Biol Chem 264: 6376–6380, 1989 King MJ, Sharma RK: N-myristoyl transferase assay using phosphocellulose paper binding. Anal Biochem 199: 149–153, 1991 Pillia S, Baltimore D: Myristoylation and the post-translational acquisition of hydrophobicity by the membrane immunoglobulin heavy-chain polypeptide in B lymphocytes. Proc Natl Acad Sci USA 84: 7654–7658, 1987 Vijayasarathy C, Bhat NR, Avadhani NG: Intramitochondrial fatty acylation of a cytoplasmic imported protein in animal cells. J Biol Chem 264: 7772–7775, 1989 da Silva AM, Klein C: Rapid posttranslational myristoylation of a 68 kDa protein in D. discoidedum. J Cell Biol 111: 401–407, 1990 Manenti S, Sorokine O, Dorsselaer AV, Taniguchi H: Demyristoylation of the major substrate of protein kinase C (MARCKS) by the cytoplasmic fraction of brain synaptosomes. J Biol Chem 269: 8309– 8313, 1994 Towler DA, Eubanks SR, Towery DS, Adams SP, Glaser L: Aminoterminal processing of proteins by N-myristoylation: Substrate specificity of N-myristoyl transferase. J Biol Chem 262: 1030–1036, 1987 Towler D, Glaser L: Protein fatty acid acylation: Enzymatic synthesis of an N-myristoyl-glycyl peptide. Proc Natl Acad Sci USA 83: 2812– 2816, 1986 Towler DA, Adams SP, Eubanks SR, Towery DS, Jackson-Machelski E, Glaser L, Gordon JI: Purification and characterization of myristoylCoA:protein N-myristoyltransferase. Proc Natl Acad Sci USA 84: 2708–2712, 1987 Towler DA, Adams SP, Eubanks SR, Towery DS, Jackson-Machelski E, Glaser L, Gordon JI: Myristoyl CoA: protein N-myristoyltransferase activities from rat liver and yeast possess overlapping yet distinct peptide substrate specificities. J Biol Chem 263: 1784–1790, 1988 Heuckeroth RO, Towler DA, Adams SP, Glaser L, Gordon JI: 11(Ethylthio) undecanoic acid. A myristic acid analog of altered hydrophobicity which is functional for peptide N-myristoylation with wheat germ and yeast acyltransferase. J Biol Chem 263: 2127–2133, 1988 King MJ, Sharma RK: Demonstration of multiple forms of bovine brain myristoyl CoA:protein N-myristoyltransferase. Mol Cell Biochem 133: 77–81, 1992 McIlhinney RAJ, McGlone K: Characterization of a myristoylCoA:glycylpeptide N-myristoyltransferase activity in rat brain. Subcellular and regional distribution. J Neurochem 54: 110–117, 1990 Raju RVS, Datla RSS, Sharma RK: Overexpression of human Nmyristoyltransferase utilizing a T7 polymerase gene expression system. Prot Exp Pur 7: 431–437, 1996 Raju RVS, Anderson JW, Datla RSS, Sharma RK: Molecular cloning and biochemical characterization of bovine spleen myristoylCoA:protein N-myristoyltransferase. Arch Biochem Biophys 134: 134– 142, 1997 King MJ, Sharma RK: Mechanisms of action of NIP71 on N-myristoyltransferase activity. Mol Cell Biochem 141: 79–86, 1994 McIlhinney RAJ, McGlone K, Willis AC: Purification and partial sequencing of myristoyl CoA:protein N-myristoyltransferase from bovine brain. Biochem J 290: 405–410, 1993

152 48. Glover CJ, Felsted RL: Identification and characterization of multiple forms of bovine brain N-myristoyltransferase. J Biol Chem 270: 23226–23233, 1995 49. Boutin JA, Ferry G, Renould AP, Maes P, Remound G, Vincent M: Myristoyl CoA:protein N-myristoyltransferase activity in cancer cells. Purification and characterization of a cytosolic isoform from themurine leukaemia cell line L1210. Eur J Biochem 214: 853–867, 1993 50. McIhinney RAJ, McGlove K: Immunocytochemical characterization and subcellular localization of human N-myristoylCoA:protein N-myristoyltransferase in HeLa cells. Exp Cell Res 223: 348–356, 1996 51. Duronio RJ, Reed SI, Gordon JI: Mutations of human myristoyl CoA:protein N-myristoyltransferase cause temperature-sensitive myristic acid auxotrophy in Saccharomyces cerevisiae. Proc Natl Acad Sci USA 89: 4129–4133, 1992 52. Cheng HC, Nishio H, Hatase O, Ralph S, Wang JH: A synthetic peptide derived from p34cdc2 is a specific and efficient substrate of Src-family tyrosine kinases. J Biol Chem 267: 9248–9256, 1992 53. Raju RVS, Kalra J, Sharma RK: Purification and properties of bovine spleen N-myristoyl CoA:protein N-myristoyltransferase. J Biol Chem 269: 12080–12083, 1994 54. Magnuson B, Raju RVS, Sharma RK: N-myristoyl transferase activity in rabbit intestine. Biochim Biophys Acta 1300: 119–124, 1996 55. Wice BM, Gordon JI: A strategy for isolation of cDNAs encoding proteins affecting human intestinal epithelial cell growth and differentiation: Characterization of a novel gut specific N-myristoylated annexin. J Cell Biol 116: 405–422, 1992 56. Bandorowicz J, Pikula S: Annexins-multifunctional, calcium dependent phospholipid binding proteins. Acta Biochem Pol 40: 281–289, 1993 57. Pullman WE, Bodmer WF: Cloning and characterization of a gene that regulates cell adhesion. Nature 356: 529–532, 1992 58. Iozzo RV, Kovalszky I, Hacobian N, Schick PK, Ellingson JS, Dodge GR: Fatty acylation of heparan sulfate proteoglycan from human colon carcinoma cells. J Biol Chem 265: 1980–1998, 1990 59. Cartwright CA, Mammajiwala S, Skolnick SA, Eckhart W, Burgess DR: Intestinal crypt cells contain higher levels of cytoskeletal-associated pp60c-src protein tyrosine kinase activity than do differentiated enterocytes. Oncogene 8: 1033–1039, 1993 60. Zhao YH, Krueger JG, Sudol M: Expression of cellular yes protein in mammalian tissues. Oncogene 5: 1629–1635, 1990 61. Kitamato Y, Yuan X, Wu Q, McCourt DW, Sadler JE: Enterokinase the initiator of intestinal digestion is a mosaic protease composed of a distinctive assortment of domains. Proc Natl Acad Sci USA 91: 7588– 7592, 1994 62. Raju RVS, Kakkar R, Datla RSS, Radhi JM, Sharma RK: MyristoylCoA:protein N-myristoyltransferase from bovine cardiac muscle. Molecular cloning, kinetic analysis, and in vitro proteolytic cleavage by m-calpain. Exp Cell Res 241: 23–35, 1998 63. McIlhinney RAJ, McGlone K: A simplified assay for the enzyme responsible for the attachment of myristic acid to the N-terminal glycine residue of proteins, myristoyl-CoA:glycylpeptide Nmyristoyltransferase. Biochem J 263: 387–391, 1989 64. Glover CJ, Goddard C, Felsted RL: N-myristoylation of pp60src. Biochem J 250: 485–491, 1988 65. Rocque WJ, McWherter CA, Wood DC, Gordon, JI: A comparative analysis of the kinetic mechanism and peptide substrate specificity of human and Saccharomyces cerevisiae myristoyl-CoA:protein Nmyristoyltransferase. J Biol Chem 268: 9964–9971, 1993 66. McIlhinney RAJ, Patel PB, McGlone K: Characterization of polyhistidine-tagged form of human myristoyl-CoA:protein N-myristoyltransferase produced in Escherichia coli. Eur J Biochem 222: 137–146, 1994

67. Raju RVS, Datla RSS, Sharma RK: Expression of human Nmyristoyltransferase in Escherichia coli. Comparison with Nmyristoyltransferases expressed in different tissues. Mol Cell Biochem 155: 69–76, 1996 68. Zhang L, Jackson-Machelski E, Gordon JI: Biochemical studies of Saccharomyces cerevisiae myristoyl-CoA:protein N-myristoyltransferase mutants. J Biol Chem 271: 33131–33140, 1996 69. Ntwasa M, Egerton M, Gay NJ: Sequence and expression of drosophila myristoyl-CoA:protein N-myristoyltransferase: Evidence for proteolytic processing and membrane localization. J Cell Sci 110: 149– 156, 1997 70. Giang DK, Cravatt BF: A second mammalian N-myristoyltransferase. J Biol Chem 273: 6595–6598, 1998 71. Raju RVS, Datla RSS, Kakkar R, Sharma RK: Recombinant bovine spleen myristoyl-CoA:protein N-myristoyltransferase. Mol Cell Biochem 189: 91–97, 1998 72. Bhatnagar RS, Futterer K, Farazi TA, Korolev S, Murray CL, JacksonMachelski E, Gokel GW, Gordon JI, Waksman G: Structure of Nmyristoyltransferase with bound myristoyl-CoA and peptide substrate analogs. Nat Struct Biol 5: 1091–1097, 1998 73. Rudnick DA, Johnson RL, Gordon JI: Study of catalytic activities and substrate specificities of Saccharomyces cerevisiae myristoylCoA:protein N-myristoyltransferase deletion mutants and human/yeast chimeras in Escherichia coli and S. cerevisiae. J Biol Chem 267: 23852–23861, 1992 74. Lodge JK, Johnson RL, Weinberg RA, Gordon JI: Comparison of myristoyl-CoA:protein N-myristoyltransferase from three pathogenic fungi: Cyrptococcus neoformans, Histoplasma capsulatum and Candia albicans. J Biol Chem 269: 2996–3009, 1994 75. Glover CJ, Hartman KD, Felsted RL: Human N-myristoyltransferase amino-terminal domain involved in targeting enzyme to the ribosomal subcellular fraction. J Biol Chem 272: 28680–28689, 1997 76. Pesecks SM, Resh MD: Fatty acyl transfer by human N-myristoyltransferase is dependent upon conserved cysteine and histidine residues. J Biol Chem 269: 30888–30892, 1994 77. McIlhinney RAJ, Young K, Egerton M, Camble R, White A, Soloviev M: Characterization of human and rat brain myristoyl-CoA:protein N-myristoyltransferase: Evidence for an alternative splice variant of the enzyme. Biochem J 333: 491–495, 1998 78. Raju RVS, Datla RSS, Sharma RK: Genomic organization of human N-myristoyltransferase-1. Biochem Biophys Res Commun 257: 284– 288, 1999 79. Duronio RJ, Towler DA, Heuckeroth RO, Gordon JI: Disruption of the yeast N-myristoyl transferase gene causes recessive lethality. Science 243: 796–800, 1989 80. Raju RVS, Sharma RK: Protein myristoylation in cell signalling. In: Sharma et al. (eds). Adaptation Biology and Medicine. Narosa Publishing House, New Delhi, India, 1997, pp 272–283 81. King MJ, Sharma RK: Identification, purification and characterization of a membrane-associated N-myristoyltransferase inhibitor protein from bovine brain. Biochem J 291: 635–639, 1993 82. Magnuson BA, Raju RVS, Moyana TN, Sharma RK: Increased Nmyristoyltransferase activity observed in rat and human colonic tumors. J Natl Cancer Inst 87: 1630–1635, 1995 83. King MJ, Sharma RK: Differential activation of bovine brain Nmyristoyl transferase(s) by a novel cytosolic activator. Biochem Biophys Res Commun 212: 580–588, 1995 84. Raju RVS, Sharma RK: Demonstration and purification of a myristoylCoA binding protein from cardiac muscle. Life Sci 60: 2145–2153, 1997 85. Nicito M, Roy G, Villar LM, Palacois C, Serrano A, Alvarez-Cermeno JC, Gonzalez-Porque P: Thioesterase and protein deacylase activities of porcine pancreatic phospholipase A2. Biochem Biophys Acta 1299: 17–22, 1996

153 86.Schmidt MFG: Fatty acylation of proteins. Biochem Biophys Acta 988: 411–426, 1989 87. Schmidt MFG: Fatty acid binding: A new kind of posttranslational modification of membrane proteins. Curr Trop Microbiol Immunol 102: 101–129, 1983 88. Lambrecht B, Schmidt MFG: Membrane fusion induced by influenza virus hemagglutinin requires protein bound fatty acids. FEBS Lett 202: 127–132, 1986 89. Barnes JA, Gomes AV: PEST sequences in calmodulin-binding proteins. Mol Cell Biochem 149/150: 17–27, 1995 90. Rogers S, Wells R, Rechsteiner M: Amino acid sequences common to rapidly degraded proteins: The PEST hypothesis. Science 234: 364– 369, 1986 91. Rechsteiner M, Rogers S: PEST sequences and regulation by proteolysis. Trends Biochem Sci 21: 267–271, 1996 92. Saido TC, Sorimachi S, Suzuki K: Calpains: New perspectives in molecular diversity and physiological-pathological involvement. FASEB J 8: 814–822, 1994 93. Caroall DE, Demartino GN: Calcium-activated neutral protease (calpain) system: structure, function and regulation. Physiol Rev 71: 813–447, 1991 94. Kakkar R, Raju RVS, Sharma RK: In vitro generation of an active calmodulin-independent phosphodiesterase from brain calmodulindependent phosphodiesterase (PDE1A2) by m-calpain. Arch Biochem Biophys 358: 320–328, 1998 95. Kakkar R, Raju RVS, Mellgren RL, Radhi J, Sharma RK: Cardiac high molecular weight calmodulin binding protein contains calpastatin activity. Biochemistry 36: 11550–11555, 1997 96. Hamakubo T, Kannagi R, Murachi T, Matus A: Distribution of calpains I and II in rat brain. J Neurosci 6: 3103–3111, 1986 97. Kamakura K, Ishiura S, Imajoh S, Nagata N, Sugita H: Distribution of calcium-activated neutral protease inhibitor in the central nervous system of the rat. J Neurosci Res 31: 543–548, 1992 98. Mellgren RL: Proteolysis of nuclear proteins by µ-calpain and mcalpain. J Biol Chem 266: 13920–13924, 1991 99. Raju RVS, Kakkar R, Radhi J, Sharma RK: Biological significance of phosphorylation and myristoylation in the regulation of cardiac muscle proteins. Mol Cell Biochem 176: 135–143, 1997 100. Stabach PR, Cianci CD, Glantz SB, Zhang A, Morrow JS: Site-directed mutagenesis of alpha II spectrin at codon 1175 modulates its calpain susceptibility. Biochemistry 36: 57–65, 1997 101. Ellis RE, Yuan J, Horvitz HR: Mechanisms and functions of cell death. Ann Rev Cell Biol 7: 663–698, 1991 102. White E: Life and death, and the pursuit of apoptosis. Genes Dev 10: 1–15, 1996 103. Nagata S: Apoptosis by death factor. Cell 88: 355–365, 1997 104. Chow M, Moscufo N: Myristoyl modification of viral proteins: Assays to assess functional roles. Meth Enzymol 250: 495–509, 1995 105. Bryant ML, Heuckeroth RO, Kimata TJ, Ratnar L, Gordon JI: Replication of human immunodeficiency virus 1 and Moloney murine leukemia virus is inhibited by different heteroatom-containing analogs of myristic acid. Proc Natl Acad Sci USA 86: 8655–8659, 1989 106. Harper DR, Gilbert RL, Blunt CJ, McIlhinney RAJ: Inhibition of varicella-zoster virus replication by an inhibitor of protein myristoylation. J Gen Virol 74: 1181–1184, 1993 107. Weinberg RA, McWherter CA, Freeman SK, Wood DC, Gordon JI: Genetic studies reveal that myristoylCoA:protein N-myristoyltransferase is an essential enzyme in Candida albicans. Mol Microbiol 16: 241– 250, 1995 108. Lodge JK, Jackson-Machelski E, Toffaletti DL, Perfect JR, Gordon JI: Targeted gene replacement demonstrates that myristoyl-CoA:protein N-myristoyltransferase is essential for viability of Cryptococcus neoformans. Proc Natl Acad Sci USA 91: 12008–12012, 1994

109.Devadas B, Freeman SK, Zupec ME, Lu HF, Nagarajan SR, Kishore NS, Lodge JK, Kuneman DW, McWherter CA, Vinjamoori DV, Getman DP, Gordon JI: Design and synthesis of novel imidazole-substituted dipeptide amides as potent and selective inhibitors of Candida albicans myristoylCoA:protein Nmyristoyltransferase and identification of related tripeptide inhibitors with mechanism-based antifungal activity. J Med Chem 40: 2609–2615, 1997 110. Raju RVS, Moyana TN, Sharma RK: N-myristoyltransferase overexpression in human colorectal adenocarcinomas. Exp Cell Res 235: 145–154, 1997 111. Felsted RL, Glover CJ, Hartman K: Protein myristoylation as a chemotherapeutic target for cancer. J Natl Cancer Inst 87: 1571–1573, 1995 112. Nadler MJS, Harrison ML, Ashendel CL, Cassady JM, Geahlen RL: Treatment of T cells with 2-hydroxymyristic acid inhibits the myristoylation and alters the stability of p56lck. Biochemistry 32: 9250–9255, 1993 113. Farschon DM, Couture C, Mustelin T, Newmeyer DD: Temporal phase in apoptosis defined by the action of Src homology 2 domains, ceramide, Bcl-2, interleukin-1beta converting enzyme family proteases, and a dense membrane fraction. J Cell Biol 137: 1117– 1125, 1997 114. Gao WD, Liu Y, Mellgren RL, Marban E: Intrinsic myofilament alterations underlying the decreased contractility of stunned myocardium. A consequence of Ca2+ dependent proteolysis? Circ Res 78: 455–465, 1996 115. Deshpande RV, Goust JM, Hogan EL, Banik NL: Calpain secreted by activated human lymphoid cells degrades myelin. J Neurosci Res 42: 259–265, 1995 116. Raju RVS, Datla RSS, Warrington RC, Sharma RK: Effects of Lhistidine and its structural analogs on human N-myristoyltransferase activity and importance of EEVEH amino acid sequence for enzyme activity. Biochemistry 37: 14928–14936, 1998 117. Weston SA, Camble R, Colls J, Rosenbrock G, Taylor I, Egerton M, Tucker AD, Tunnicliffe A, Mistry A, Mancia F, de la Fortelle E, Irwin J, Bricogne G, Paupti RA: Crystal structure of anti-fungal target Nmyristoyl transferase. Nat Struct Biol 5: 213–221, 1998 118. Warrington RC: Modulation of anticancer drug toxicity by L-histidinol: A route for improving human cancer chemotherapy? Drugs Future 18: 743–749, 1993 119. Warrington RC, Muzyka TG, Fang WD: Histidinol-mediated improvement in the specificity of 1-beta-D-arabinofuranosylcytosine and 5-fluorouracil in L1210 leukemia-bearing mice. Cancer Res 44: 2929– 2935, 1984 120. Warrington RC, Fang WD: Histidinol-mediated enhancement of the specificity of two anticancer drugs in mice bearing leukemia bone marrow disease. J Natl Cancer Inst 74: 1071–1077, 1985 121.Warrington RC, Cheng I, Fang WD: Effects of L-histidinol on the susceptibility of P815 mastocytoma cells to selected anticancer drugs in vitro and in DBA/2J mice. J Natl Cancer Inst 78: 1177–1183, 1987 122.Warrington, RC, Fang, WD: L-histidinol improves the selectivity and efficacy of alkylating agents and daunomycin in mice with P388 leukaemia. Br J Cancer 60: 652–656, 1989 123. Warrington RC, Fang WD: Improved treatment of disseminated B16f10 melanoma in mice with anticancer drugs in combination with Lhistidinol. Anticancer Res 11: 1869–1874, 1991 124. Warrington RC, Fang WD, Zhang LU: L-histidinol reverses resistance to cisplatinum and other antineoplastics in tumorigenic epithelial cell line. Anticancer Res 16: 3641–3646, 1996 125. Krueger JG, Garber AE, Goldberg AR, Hanafusa H: Changes in aminoterminal sequences of pp60src lead to decreased membrane association and decreased in vivo tumorigenicity. Cell 28: 889–896, 1982

154 126.Cross FR, Garber EA, Pellman D, Hanafusa HA: A short sequence in the p60src N-terminus is required for p60src myristoylation and membrane association and for cell transformation. Mol Cell Biol 4: 1834–1842, 1984 127. Kamps MP, Buss JE, Sefton BM: Rous sarcoma virus transforming protein lacking myristic acid phosphorylates known polypeptides substrates without inducing transformation. Cell 45: 105–112, 1986 128. Shoji S, Kurosawa T, Inoue H, Funakoshi T, Kubota Y: Human cellular src gene product: identification of the myristoylated pp60src and blockage of its myristoyl acylation with N-fatty acyl compounds resulted in the suppression of colony formation. Biochem Biophys Res Commun 173: 894–901, 1990 129. Blenis J, Resh MD: Subcellular localization specified by protein acylation and phosphorylation. Curr Opin Cell Biol 5: 984–989, 1993 130. Wedegaertner PB, Wilson PT, Bourne HR: Lipid modifications of trimeric G proteins. J Biol Chem 270: 503–506, 1995 131. Toribara NW, Ho SB, Bresalier RS, Kim YS: Biochemical changes in colorectal carcinogenesis. In: H.K. Seitz, U.A. Simanowski (eds). Colorectal Cancer: From Pathogenesis to Prevention. Springer Verlag, Berlin, 1985, pp 256–288 132. Haggitt RC, Glotzbach RE, Soffer EE, Wruble LD: Prognostic factors in colorectal caricinomas arising in adenomas: Implications for lesions removed by endoscopic polypectomy. Gastroenterology 89: 328–336, 1985 133. Galandiuk S, Fazio VW, Jagelman DG, Lavery IC, Weakley FA, Petras RE, Badhwar K, McGonagle B, Eastin K, Sutton T: Villous and tubulovillous adenomas of the colon and rectum. (A retrospective review, 1964–1985). Am J Surg 153: 41–47, 1987 134. Harlow SP, Ericksen BL, Poggensee L, Chmiel JS, Scarpelli DG, Murad T, Bauer KD: Prognostic implications of proliferative activity and DNA aneuploidy in Astler-Coller Dukes stage C colonic adenocarcinomas. Cancer Res 51: 2403–2409, 1991 135. Steinberg SM, Barkin JS, Kaplan RS, Stablein DM: Prognostic indicators of colon tumors. The Gastrointestinal Tumor Study Group experience. Cancer 57: 1866–1870, 1986 136. Lev R, Lance MP: Adenomatous Polyps of the Colon: Pathobiological and Clinical Features. Springer-Verlag, New York, 1990 137. Aderem AA, Keum MM, Pure E, Cohn ZA: Bacterial lipopolysaccharides, phorbol myristate acetate, and zymosan induce the myristoylation of specific macrophage proteins. Proc Natl Acad Sci USA 83: 5817–5821, 1986 138. Polio G, Sorio C, Berton G: Protein myristoylation in human mononuclear phagocytes: Modulation by interferon-γ and tumor necrosis factor-α. J Cell Sci 100: 833–840, 1991 139. Bolen JB, Veillett A, Schwartz AM, DeSeau V, Rosen N: Activation of pp60c-src protein kinase activity in human colon carcinoma. Proc Natl Acad Sci USA 84: 2251–2255, 1987 140. Cartwright CA, Kamps MP, Meiser AI, Pipas JM, Eckhart W: pp60c-src activation in human colon carcinoma. J Clin Invest 83: 2025–2033, 1989 141. Park J, Meisler AI, Cartwright CA: c-Yes tyrosine kinase activity in human colon carcinoma. Oncogene 8: 2627–2635, 1993 142. Hamaguchi M, Hanafusa H: Association of p60src with Triton S-100 resistant cellular structure correlates with morphological transformations. Proc Natl Acad Sci USA 84: 2312–2316, 1987 143. Buss JE, Solski PA, Schaeffer JP, MacDonald MJ, Der CJ: Activation of cellular p21 ras by myristoylation. Biochem Soc Trans 17: 867– 869, 1989 144. Cadwallader KA, Paterson H, MacDonald SG, Hancock JF: Nterminally myristoylated Ras proteins require palmitoylation or a polybasic domain for plasma membrane localization. Mol Cell Biol 14: 4722–4730, 1994

145.Raju RVS, Moyana TN, Datla RSS, Kakkar R, Sharma RK: Nmyristoyltransferase (NMT) as a marker for diagnosis and prognosis of colon cancer. In: P. Smith, P. Warne (eds). Emerging Therapeutic Targets, vol. 2. Ashley Publications, London, UK, 1998, pp 129–131 146. Raju RVS, Datla RSS, Sharma RK: Reduction of oncoprotein transformation in vitro by albumin. J Natl Cancer Inst 88: 556–557, 1996 147. Nagase S, Shimamune K, Shumiya S: Albumin-deficient rat mutant. Science 205: 590–591, 1979 148. Kakizoe T, Sugimura T: Chemical carcinogenesis in analbuminmic rats. Jpn J Cancer Res 79: 775–784, 1988 149. King MJ, Pugazhenthi S, Khandelwal RL, Sharma RK: Elevated Nmyristoyl transferase activity is reversed by sodium orthovanadate in streptozotocin-induced diabetic rat. Biochem Biophys Acta 1165: 259– 262, 1993 150. Rothschild MA, Oratz M, Schreiber SS: Albumin synthesis I. N Eng J Med 286: 748–757, 1997 151. Era S, Kuwata K, Imai H, Nakamura K, Hayashi Y, Sogami M: Age related changes in redox state of human serum albumin. Biochem Biophys Acta 1247: 12–16, 1995 152. Majumdar AP, Tureaud J, Relan NK, Kessel A, Dutta S, Hatifiled JS, Fligiel SE: Increased expression of pp60c-src in gastric mucosa of aged rats. J Gerontol 49: B110–B116, 1994 153. Pugazhenthi S, Khandelwal RL, Angel JF: Insulin-like effect of vanadate on malic enzyme and glucose-6-phosphate dehydrogenase activities in streptozotocin-induced diabetic rat liver. Biochem Biophys Acta 1083: 310–312, 1991 154. Pugazhenthi S, Angel JF, Khandelwal RL: Long term effects of vanadate treatment metabolizing and lipogenic enzymes of liver in genetically diabetic (db/db) mice. Metabolism 40: 941 –946, 1991 155. Khandelwal RL, Pugazhenthi S: In vivo effects of vanadate on hepatic glycogen metabolism and lipogenic enzymes in insulin-dependent and insulin-resistant diabetic animals. Mol Cell Biochem 153: 87– 94, 1995 156. King MJ, Sharma RK: Membrane associated N-myristoyltransferase activity is reduced in obese (fa/fa) Zucker rat liver. Biochem Biophys Res Commun 197: 665–670, 1993 157. King MJ, Pugazhenthi S, Khandelwal RL, Sharma RK: In vivo modulation of N-myristoyltransferase activity by orthovanadate. Mol Cell Biochem 153: 151–153, 1995 158. Remy R, Tremolieres A, Duval JC, Ambard-Bretteville F, Dubacq JP: Study of the supramolecular organization of light-harvesting chlorophyll protein (LHCP). Conversion of the oligomeric form into the monomeric one by phospholipase A2 and reconstitution with liposomes. FEBS Lett 137: 271–275, 1982 159. Mattoo AK, Edelman M: Intramembrane translocation and posttranslational palmitoylation of the chloroplast 32 kDa herbicidebinding protein. Proc Natl Acad Sci USA 84: 1497–1501, 1987 160. Quinn PJ, Williams WP: The structural role of lipids in photosynthetic membranes. Biochem Biophys Acta 737: 223–266, 1983 161. Hirayama T, Oka A: Novel protein kinase of Arabidopsis thaliana (APK1) that phosphorylates tyrosine, serine and threonine. Plant Mol Biol 20: 653–662, 1992 162. Loh YS, Zhou J, Martin GB: The myristoylation motif of Pto is not required for disease resistance. Mol Plant Microbe Interact 11: 572– 576, 1998 163. Martin GB, Frary A, Wu T, Brommonschenkel S, Chunwongse J, Earle ED, Tanksley SD: A member of the Pto gene family confers sensitivity to fenthion resulting in rapid cell death. Plant Cell 6: 1543–1552, 1994 164. Saijo Y, Hata S, Sheen J, Izui K: A cDNA cloning and prokaryotic expression of maize calcium-dependent protein kinase. Biochem Biophys Acta 1350: 109–114, 1997

155 165.Tregear JW, Jouannic S, Schwebel-Dugue N, Kreis M: An unusual protein kinase displaying characteristics of both the serine/ threonine and tyrosine families is encoded by the Arabidopsis thaliana gene ATN1. Plant Sci 117: 107–119, 1996 166.Fulda M, Heinz E, Wolter F: Brassica napus cDNAs encoding

fatty acyl-CoA synthetase. Plant Mol Biol 33: 911–922, 1997 167.Pawlowski K, Twigg P, Dobritsa S, Guan C, Mullin BC: A nodulespecific gene family from Alnus glutinosa encodes glycine-and histidine-rich proteins expressed in the early stages of actinorhizal nodule development. Mol Plant Microbe Interact 10: 656–664, 1997