BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS ARTICLE NO.
225, 1045–1051 (1996)
1292
Human Dermal Fibroblasts Express Eotaxin: Molecular Cloning, mRNA Expression, and Identification of Eotaxin Sequence Variants Joachim Bartels,*,1 Carsten Schlu¨ter,* Elvira Richter,† Norio Noso,* Reinhard Kulke,* Enno Christophers,* and Jens-M. Schro¨der* *Clinical Research Unit, Department of Dermatology, University of Kiel, D-24105 Kiel, Germany; and †Department of Immunology and Cell Biology, Forschungsinstitut Borstel, D-23845 Borstel, Germany Received July 15, 1996 Recently we discovered and purified a novel b-chemokine with eosinophil specific chemotactic activity from supernatants of long-term TNF-a stimulated dermal fibroblasts. Using degenerated specific oligonucleotides based on partial amino acid sequence data and a PCR protocol, we obtained different clones sharing high sequence homology with this novel chemokine and with human eotaxin cDNA. Semi-quantitative RTPCR experiments using eotaxin-specific primers indicate low constitutive eotaxin mRNA expression in human dermal fibroblasts which is upregulated by IL-1a and TNF-a within 6 hrs and modulated by IFNg. While IL-1a-induced eotaxin mRNA accumulation is transient, long-term stimulation with TNF-a resulted in a further increase of eotaxin mRNA. q 1996 Academic Press, Inc.
Affected skin of patients suffering from atopic dermatitis shows deposits of eosinophil (Eo) derived proteins. Occasionally intact Eos are present as one of the characteristic features of this disease (1). The function of Eos in the patho-physiology of this skin disease and other diseases including late phase allergic reactions is still incompletely understood (2). Disruption of Eos and their degranulation causes severe tissue destruction and therefore may be important in the inflammatory process underlying these diseases. The appearance of Eo-deposits in the dermis but not epidermis of affected skin of atopic patients led to the working hypothesis that dermal cells, i.e. fibroblasts, generate Eo-attractants. Recently we discovered and purified a novel b-chemokine from supernatants of long-term TNF-a stimulated dermal fibroblasts which we named ‘‘eotactin’’ because of its eosinophil specific chemotactic activity (3,4). To isolate and clone the corresponding cDNA from dermal fibroblasts we used degenerated oligonucleotides based on partial peptide sequence data and a RACE (rapid amplification of cDNA ends, (5)) protocol. In this paper we report as one result of this approach the isolation, characterisation and expression of human eotaxin cDNA in dermal fibroblasts as well as the identification of eotaxin cDNA variants. MATERIALS AND METHODS 1. Isolation, culture and stimulation of dermal fibroblasts. Primary dermal fibroblasts were prepared from foreskin and cultured as previously reported (6). Dermal fibroblasts were stimulated with 20 ng/ml TNF-a in serum free growth medium for 96 hrs unless otherwise stated. For PCR gene expression studies fibroblasts were cultured in DMEM lacking FCS for 24 hrs prior to stimulation to exclude FCS mediated changes in gene expression. Recombinant stimuli used were from Pepro Tech Inc., London, UK. 2. Oligonucleotide design based on partial protein sequence data. Production, purification, chemotactic assays and protein sequencing of eotactin were done as described (4). Two degenerated anti-sense primers for use in nested 5*RACE reactions were designed as depicted in Figure 1: primer ETC-LG was designed for initial cDNA amplification
1
To whom correspondence should be addressed. Fax: (/49)-431-597-1611. E-mail:
[email protected]. 1045 0006-291X/96 $18.00 Copyright q 1996 by Academic Press, Inc. All rights of reproduction in any form reserved.
AID
BBRC 5217
/
6907$$$141
08-06-96 13:54:48
bbrca
AP: BBRC
Vol. 225, No. 3, 1996
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
of eotactin and related CC-Chemokines and primer ETC-LN-BAMH1 was designed as eotactin specific primer for subsequent nested PCR reactions. 3. cDNA cloning and sequencing. Total RNA from stimulated dermal fibroblasts was isolated (7) and reverse transcribed using standard reagents (GIBCO/BRL, Eggenstein, FRG) and a primer (5*-CCTGTCGACGGTACCAAGCUTTUTUTUTTTTTTT-3*) composed of an anchor-sequence (underlined) for cloning and PCR purposes attached to an UT rich part designed to anneal to poly A tails of mRNA molecules. 5* RACE was done following an inverse PCR protocol (8) taking advantage of the fact that the known sequence of the primer used for generating single stranded cDNA becomes attached to the 5*-end of the desired cDNA after second strand cDNA synthesis and ligation of the ds-cDNA products. In initial rounds of amplification (37 cycles) ETC-LG was used as gene specific primer and FR-inv-S (5*-GCTTGGTACCGTCGACAG-3*) as forward primer, complementary to the incor porated anchorsequence of the first strand cDNA primer. PCR products at this and later stages (as well as recombinant clones) were screened using a nonradioactive detection protocol (9) after low stringency (607C) southern hybridization with a MCP3 cDNA probe: A major PCR product of approx. 320 bps was isolated and served as template for the subsequent nested PCR reaction (37 cycles) using FR-inv as anchor primer (same as FR-inv-S but with an additional G nucleotide at its 3*-end in order to serve as nested primer) and ETC-LN-BAMH1 as gene specific nested primer. The PCR products (approx. 280 bps) were cloned into a pZErO vector (Invitrogen, The Netherlands). For 3*-RACE a sense primer (EOF3: 5*-CCAMYTCTCACGCCAAAGCTCACAC-3*; M Å A or C; Y Å C or T) based on the sequence of eotaxin encoding 5*-RACE products (clones 4, 10, 11, 14, 15) was used in conjunction with the anchor primer FR-B (5*-GCTTGGTACCGTCGACAG-3*). PCR products of approx. 780 bps were directly cloned into a pTAg vector (R&D Systems GmbH, Wiesbaden, FRG) and sequenced. Plasmid nucleotide sequences were determined with the chain termination method (10) using an automated laser fluorescent DNA sequencer (ALF, Pharmacia, Freiburg, FRG) or an ABI PRISM Model 377 sequencer (Perkin-Elmer/Applied Biosystems, Weiterstadt, FRG). 4. Semi-quantitative duplex RT-PCR (SQRT-PCR). Intron spanning sets of primers specific for GAPDH (11) and eotaxin (see Fig. 2; forward primer: 5*-CCCAACCACCTGCTGCTTTAACCTG-3*, reverse primer: 5*-TGGCTTTGGAGTTGGAGATTTTTGG-3*) were designed to differentiate between genomic and cDNA-templates. 1 mg total RNA was reverse transcribed using standard reagents (GIBCO/BRL, Eggenstein, FRG). cDNA corresponding to 50 ng RNA served as template in a duplex-PCR-reaction containing 0,8 mM of eotaxin specific primers and (as internal control for equal amounts of cDNA before PCR) 0,1 mM of a GAPDH specific primer pair (11,12). Amplification was done using 30 cycles with denaturation at 947C for 45 sec., primer annealing at 607C for 20 sec. and primer extension at 727C for 30 sec. initially increasing by 3 seconds after each cycle. The PCR products were size-fractionated by agarose gelelectrophoresis and visualised by ethidium bromide staining. The results were quantified by scanning Polaroid pictures using a densitometer (Desaga CD60). Specificity of eotaxin encoding PCR products were verified by the presence of an eotaxin characteristic EcoRV restriction site (Fig. 4).
RESULTS
cDNA Cloning and Sequence Analysis of Eotaxin and Related Sequences Amino acid sequences of internal peptide fragments derived from the novel chemokine were used to design degenerated oligonucleotides (Figure 1) suitable for nested 5*-RACE. After improvement of annealing conditions in 5*-RACE we were able to obtain a cDNA clone with close homology to MCP-4 ((13), manuscript in preparation) and several other approx. 280 bps long 5*-cDNA clones sharing high sequence homology with human MCP-chemokines and rodent eotaxins as well as with the peptide fragments obtained from the new chemokine. While there was no matching cDNA sequence accessible in the GenBank/EMBL sequence databases at the time of our sequence submission, recent publications (14-16) show, that almost all of these cDNA clones are identical with human eotaxin cDNA; one 5*-cDNA clone contained 4 deviations in the nucleotide sequence resulting in 2 amino acid changes (clone 34, Figure 3). Based on the eotaxin 5*-cDNA sequence a sequence specific primer (EOF3) was designed to amplify almost the entire eotaxin cDNA in a 3*-RACE reaction. Clones derived from this experiment were identified to encode an eotaxin-variant (clone 53, see Fig. 2) differing in several nucleotides (2 within the protein coding region resulting in 2 amino acid changes, (Figure 3)) and in the length of the 3* untranslated region (UTR) from published sequences. Eotaxin mRNA Expression in Dermal Fibroblasts Eotaxin cDNA amplification relative to GAPDH cDNA amplification was used to monitor stimulus dependent changes in eotaxin mRNA expression, assuming, that mRNA expression 1046
AID
BBRC 5217
/
6907$$$141
08-06-96 13:54:48
bbrca
AP: BBRC
Vol. 225, No. 3, 1996
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
FIG. 1. Oligonucleotide design based on partial protein sequence data. (A) Part of the amino acid alignment of human MCP sequences relevant for primer construction. Conserved residues for the entire CC-chemokine family are marked by asterisks; residues common to all known human MCP’s and to eotactin are boxed. (B) Several batches of eotactin were digested with endoproteinase Asp-N or Lys-C to completion; the resulting peptides were separated by RP-HPLC and sequenced as described (4). The compiled peptide sequence of eotactin is indicated and the amino acid sequences used for oligonucleotide design are underlined. (C) Three-letter code of the peptides (see B) and design of the degenerated PCR-primers for cDNA cloning. The nested reverse primer ETC-LN-BAMH1 contains a restriction endonuclease site for cloning purposes(underlined); (I stands for Inosine). (D) cDNA sequences corresponding to the primer sequences determined after cloning.
of the housekeeping gene GAPDH is not affected by the stimuli used (12). Experiments performed with fibroblasts cultured for 24 hrs in DMEM without FCS showed only low eotaxin mRNA expression without further stimulation (Fig. 4). Addition of IL-1a or TNF-a resulted in a slight upregulation of eotaxin mRNA expression within 6 hrs of stimulation (Fig.4). While IL-1a-induced eotaxin mRNA accumulation was transient, long-term stimulation with TNFa resulted in a strong increase of eotaxin mRNA in dermal fibroblasts. IFN-g by itself had no upregulating effect but in combination with IL-1a or TNF-a further enhanced eotaxin mRNA expression within 24 hours. This synergistic effect was absent after 48 hrs of costimulation (Fig. 4). DISCUSSION
We here describe the identification of human eotaxin expression in dermal fibroblasts by cDNA cloning and SQRT-PCR: Sequence analysis of cDNA clones revealed the cysteinepattern typical for members of the b-chemokine family as well as features characteristic for eotaxin: the 2-amino acid gap at position 6 and the group of 3 positively charged amino acids at consecutive positions starting at position 56 (Lys-Lys-Lys in previously described eotaxins 1047
AID
BBRC 5217
/
6907$$$141
08-06-96 13:54:48
bbrca
AP: BBRC
Vol. 225, No. 3, 1996
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
FIG. 2. cDNA sequence and deduced protein sequence of a human eotaxin variant. Nucleotides from position 055 to 014 are from clone 4, all others from clone 53. The dotted line marks the position of primer EOF3 used to generate clone 53. Within the 3*-end untranslated region (UTR) sequence motifs indicative of short-lived mRNA species (30) are double underlined. The two polyadenylation signal sequences (21) are single underlined. The predicted open reading frame is indicated below the nucleic acid sequence. In the protein sequence, the start methionine is in uppercase, the probable signal peptide is in italic, and the mature protein is in bold letters. The arrow indicates the predicted site for a signal peptidase cleavage. The stop codon is indicated by an asterisk. The positions of the primers (r, R) used for SQRT-PCR as well as the characteristic EcoRV restriction site within the amplification product are indicated. The nucleotide and deduced amino acid sequences shown above as well as from clones 4 and 34 are available from DDBJ/GenBank/EMBL under the accession numbers Z75668, Z69291, and Z75669, respectively.
(14-19) and Lys-Lys-Arg in clone 53 described here; (Figure 3)). cDNA cloning also revealed some sequence microheterogeneity which partially effects the deduced protein sequence (Figure 3). This observation may explain at least in part the heterogeneity observed using mass spectroscopy of the purified protein (4) although nucleotide alteration resulting from misincorporation by Taq polymerase cannot completely be ruled out. Since we pooled mRNA obtained from fibroblasts of different donors for our studies, an eotaxin sequence polymorphism may explain our observations. Alternatively our results may indicate that there are closely related but different eotaxin genes. The existence of multiple closely related sequence variants has been documented for other chemokines like MIP-1 (for review see (20)) and further supports the impression, that this is a more general phenomenon among chemokines. The functional relevance of this fact is not yet clear. Two potential sites of polyadenylation have been predicted for eotaxin mRNA (15) and both seem to be functional: compared to the eotaxin cDNA derived from a human small 1048
AID
BBRC 5217
/
6907$$$141
08-06-96 13:54:48
bbrca
AP: BBRC
FIG. 3. Comparison of amino acid sequences deduced from cDNA clones of human eotaxin ((14-16) and our data) with amino acid sequences of murine (17) and guinea pig eotaxin (18,19) and those of human MCP-1 (24,25), MCP-2 (26,27), MCP-3 (28,29) and MCP-4 (13). The CC-chemokines were aligned according to the conserved cysteine residues. Amino acids aligned to position 1 represent in most cases the predicted first amino acid of the mature protein. Eotaxin amino acid residues common to all sequences of this selection are displayed in white letters on a black background. Amino acid residues identical to the human eotaxin sequence shown in the first line are represented by dashes. Gaps introduced through alignment are indicated by dots. Sequence motifs characteristic for eotaxins are boxed and shaded.
Vol. 225, No. 3, 1996
AID
BBRC 5217
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
1049
/
6907$$5217
08-06-96 13:54:48
bbrca
AP: BBRC
Vol. 225, No. 3, 1996
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
FIG. 4. Gene expression of Eotaxin relative to GAPDH in cultured human dermal fibroblasts treated with cytokines. RNA was isolated from human dermal fibroblasts treated for 0, 6, 24 and 48 hours with the indicated cytokines (TNFa: 30 ng/ml, IL-1a: 3 ng/ml, IFN-g: 20 ng/ml). SQRT-PCR was carried out starting from cDNA corresponding to 50 ng of total RNA. GAPDH (318 bp) and Eotaxin (207 bp) specific amplification products are indicated. Restriction enzyme digestion with EcoRV (‘‘/’’) was used to prove the identity of the PCR product (‘‘0’’) generated with the Eotaxin specific primers. Note that the GAPDH PCR product remains uncleaved. A 100 bp ladder was used as molecular weight size marker (MW). Top: Eotaxin mRNA expression relative to GAPDH expression was calculated using the densitometric determined amounts of the corresponding amplification products and setting the value of eotaxin/GAPDH-expression obtained under conditions of no stimulation to 1.
intestine cDNA library (16), we find a slightly longer (18 bp) 3*-UTR with a second uncommon polyadenylation signal (CATAAA, (21)) located 26 bp downstream of the uncommon ATTAAA signal site common to both sequences. The use of alternative polyadenylation sites for tissue specific transcription has been documented (22) and relates to differences in gene expression (23). That also may apply in this context for eotaxin gene expression. We have shown that the proinflammatory cytokines IL-1a and TNF-a upregulate and that IFN-g modulates eotaxin gene expression in dermal fibroblasts in a similar manner as has been shown for endothelial and respiratory epithelial cells (15). The unusual further increase 1050
AID
BBRC 5217
/
6907$$$141
08-06-96 13:54:48
bbrca
AP: BBRC
Vol. 225, No. 3, 1996
BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS
in eotaxin mRNA expression after prolonged stimulation with TNF-a however has not yet been described before for other chemokines. Eotaxin expression by dermal fibroblasts as well as by cells of the gastrointestinal and respiratory tract (14,15) may help to explain how eosinophils are recruited from the blood into tissues with an epithelial interface with the environment. Further studies need to focus on the different mechanisms of eotaxin gene expression induced by different stimuli and their role in inflammatory skin diseases. ACKNOWLEDGMENTS This work was supported by Deutsche Forschungsgemeinschaft Grant CH38/7-1. We are gratefully indebted to C. Mehrens and C. Wohlenberg for technical assistance.
REFERENCES 1. Kapp, A. (1993) Hautarzt. 44, 432–436. 2. Morsey, M. A., Popowych, Y., Kowalski, J., Gerlach, G., Godson, D., Campos, M., and Babiuk, L. A. (1996) Microb. Pathog. 20, 203–212. 3. Noso, N., Bartels, J., Christophers, E., and Schro¨der, J.-M. (1996) Arch. Dermatol. Res. 288, 326. [Abstract] 4. Noso, N., Bartels, J., von der Lage, P., Zeng Min, Christophers, E., and Schro¨der, J.-M. (1996) In preparation. 5. Frohman, M. A., Dush, M. K., and Martin, G. R. (1988) Proc. Natl. Acad. Sci. USA 85, 8998–9002. 6. Schro¨der, J.-M., Sticherling, M., Henneicke, H. H., Preissner, W. C., and Christophers, E. (1990) J. Immunol. 144, 2223–2232. 7. Chomczynski, P., and Sacchi, N. (1987) Anal. Biochem. 162, 156–159. 8. Uematsu, Y. (1991) Immunogenetics 34, 174–178. 9. Engler-Blum, G., Meier, M., Frank, J., and Mu¨ller, G. A. (1993) Anal. Biochem. 210, 235–244. 10. Sanger, F., Nicklen, S., and Coulson, A. R. (1977) Proc. Natl. Acad. Sci. USA 74, 5463–5467. 11. Schlu¨ter, C., Herrmann, F., Bartels, J., Kulke, R., Christophers, E., and Schro¨der, J.-M. (1996) In preparation. 12. Aschoff, J. M., Lazarus, D., Fanburg, B. L., and Lanzillo, J. J. (1994) Anal. Biochem. 219, 218–223. 13. Uguccioni, M., Loetscher, P., Forssmann, U., Dewald, B., Li, H. D., Lima, S. H., Li, Y. L., Kreider, B., Garotta, G., Thelen, M., and Baggiolini, M. (1996) J. Exp. Med. 183, 2379–2384. 14. Ponath, P. D., Qin, S. X., Ringler, D. J., Clark-Lewis, I., Wang, J., Kassam, N., Smith, H., Shi, X. J., Gonzalo, J. A., Newman, W., Gutierrez-Ramos, J. C., and Mackay, C. R. (1996) J. Clin. Invest. 97, 604–612. 15. Garcia-Zepeda, E. A., Rothenberg, M. E., Ownbey, R. T., Celestin, J., Leder, P., and Luster, A. D. (1996) Nature Med. 2, 449–456. 16. Kitaura, M., Nakajima, T., Imai, T., Harada, S., Combadiere, C., Tiffany, H. L., Murphy, P. M., and Yoshie, O. (1996) J. Biol. Chem. 271, 7725–7730. 17. Rothenberg, M. E., Luster, A. D., and Leder, P. (1995) Proc. Natl. Acad. Sci. USA 92, 8960–8964. 18. Jose, P. J., Adcock, I. M., Griffiths Johnson, D. A., Berkman, N., Wells, T. N., Williams, T. J., and Power, C. A. (1994) Biochem. Biophys. Res. Commun. 205, 788–794. 19. Jose, P. J., Griffiths Johnson, D. A., Collins, P. D., Walsh, D. T., Moqbel, R., Totty, N. F., Truong, O., Hsuan, J. J., and Williams, T. J. (1994) J. Exp. Med. 179, 881–887. 20. Schall, T. J. (1991) Cytokine 3, 165–183. 21. Wickens, M., and Stephenson, P. (1984) Science 226, 1045–1051. 22. Thekkumkara, T. J., Livingston, W., Kumar, R. S., and Sen, G. C. (1992) Nucleic Acids. Res. 20, 683–687. 23. Yang, H., and Melera, P. W. (1994) Nucleic Acids Res. 22, 2694–2702. 24. Furutani, Y., Nomura, H., Notake, M., Oyamada, Y., Fukui, T., Yamada, M., Larsen, C. G., Oppenheim, J. J., and Matsushima, K. (1989) Biochem. Biophys. Res. Commun. 159, 249–255. 25. Yoshimura, T., Yuhki, N., Moore, S. K., Appella, E., Lerman, M. I., and Leonard, E. J. (1989) FEBS Lett. 244, 487–493. 26. Chang, H. C., Hsu, F., Freeman, G. J., Griffin, J. D., and Reinherz, E. L. (1989) Int. Immunol. 1, 388–397. 27. Vandamme, J., Proost, P., Lenaerts, J. P., and Opdenakker, G. (1992) J. Exp. Med. 176, 59–65. 28. Opdenakker, G., Froyen, G., Fiten, P., Proost, P., and Van-Damme, J. (1993) Biochem. Biophys. Res. Commun. 191, 535–542. 29. Minty, A., Chalon, P., Guillemot, J. C., Kaghad, M., Liauzun, P., Magazin, M., Miloux, B., Minty, C., Ramond, P., Vita, N., Lupker, J., Shire, D., Ferrara, P., and Caput, D. (1993) Eur. Cytokine. Netw. 4, 99–110. 30. Chen, C. Y. A., and Shyu, A. B. (1995) Trends Biochem. Sci. 20, 465–470.
1051
AID
BBRC 5217
/
6907$$$141
08-06-96 13:54:48
bbrca
AP: BBRC
