Epidermal Interferon-γ Inducible Protein-10 (IP-10) andMonokine Induced by γ-Interferon (Mig) but not IL-8 mRNAExpression is Associated with Epidermotropism in CutaneousT Cell Lymphomas

Cornelis P. Tensen, Maarten H. Vermeer, Petra M. van der Stoop, Peter van Beek, Rik J. Scheper,Dick M. Boorsma, and Rein WillemzeDepartment of Dermatology, Free University Hospital, Amsterdam, The Netherlands

Epidermal infiltration by neoplastic CD4F T cells is acharacteristic histologic feature of early stage mycosisfungoides, the most common type of cutaneous T celllymphoma (CTCL). The mechanisms involved in epider-motropism are unknown. It has been suggested that theCXC chemokines IL-8 and interferon-γ inducible protein10 (IP-10) may play a role, but evidence that thesechemokines are produced within the epidermis in epider-motropic CTCL is lacking. In this study skin biopsiesfrom 17 CTCL patients, including 12 mycosis fungoides,four pleomorphic CTCL, and one CD8F CTCL, wereinvestigated for epidermal IL-8 and IP-10 mRNA expres-sion by RNA in situ hybridization. In addition, theexpression of monokine induced by γ-interferon (Mig)mRNA, a CXC chemokine closely related to IP-10, wasstudied as well. The expression of IL-8 receptors A andB (CXCR1 and CXCR2, respectively) was investigated

Mycosis fungoides (MF) is the most common type ofcutaneous T cell lymphoma (CTCL). Epidermalinfiltration by atypical CD31CD41CD8– T cellswith cerebriform and hyperchromatic nuclei (epid-ermotropism) is a characteristic feature in the early

patch and plaque stages of this disease. With progression to tumorstage MF the affinity of the neoplastic T cells for the epidermis mayget lost, which is often associated with a more unfavorable prognosis(Edelson, 1980). Similarly, large pleomorphic CTCL (CD30 negative),which may develop from pre-existing MF or may develop de novo inthe skin, are often nonepidermotropic and have a poor prognosis(Beljaards et al, 1994).

Recent studies have started to evaluate the mechanisms underlyingthe affinity of these neoplastic T cells for the epidermis. These studieshave focused above all on the potential role of chemokines, arecently defined superfamily of cytokines with chemotactic and immune

Manuscript received June 17, 1997; revised March 27, 1998; accepted forpublication April 3, 1998.

Reprint requests to: Dr. Rein Willemze, Department of Dermatology, FreeUniversity Hospital, De Boelelaan 1117, 1081 HV Amsterdam, The Netherlands.

Abbreviations: CXCR1, IL-8 receptor A; CXCR2, IL-8 receptor B; IP-10,interferon-γ inducible protein 10; Mig, monokine induced by γ-interferon;MF, mycosis fungoides; RISH, RNA in situ hybridization.

0022-202X/98/$10.50 · Copyright © 1998 by The Society for Investigative Dermatology, Inc.

222

by immunohistochemistry. The results were correlatedwith the number and phenotype of epidermotropic Tcells. Epidermal expression of IP-10 and Mig mRNA wasdetected in 10 of 11 and seven of 11 epidermotropicCTCL, respectively, but not in five nonepidermotropicCTCL biopsies or normal human skin. Epidermal IP-10and Mig mRNA expression correlated with epidermalinfiltration of CD4F T cells, but not of CD8F T cells.IL-8 mRNA was demonstrated in the epidermis of onlytwo of 15 CTCL biopsies, and was associated, in bothcases, with accumulation of neutrophils. Consistently,immunostaining of the (intraepidermal) T cells withantibodies against CXCR1 and CXCR2 was not observed.In conclusion, the results of this study indicate that IP-10, and to a lesser extent Mig, but not IL-8 is involvedin the preferential infiltration of neoplastic CD4F Tcells in CTCL. Key words: chemokine/CTCL/RNA in situhybridization. J Invest Dermatol 111:222–226, 1998

response modulating properties. Chemokines can be divided into foursubfamilies, C, CC, CXC, and CXXXC chemokines, depending onthe position of the characteristic cysteine residues within their aminoterminal (Miller and Krangel, 1992; Baggiolini et al, 1994; Murphy,1994; Bazan et al, 1997; Wells and Peitsch, 1997). The CC chemokines(e.g., RANTES and the MCP) mainly affect monocytes and T cells(Oppenheim et al, 1991; Miller and Krangel, 1992). The group ofCXC chemokines can be subdivided into a group containing the ELRmotif preceding the CXC sequence (e.g., IL-8 and GRO) mainlyaffecting neutrophils (Oppenheim et al, 1991), and a group which lacksthis ELR motif [e.g., interferon-γ inducible protein 10 (IP-10) andmonokine induced by γ-interferon (Mig)], now recognized to haveactivity towards activated T cells (Taub et al, 1993b; Farber, 1997). IP-10 and Mig are closely related chemokines, which share at least onereceptor termed CXCR3 (Loetscher et al, 1996). Both chemokinesare chemotactic for stimulated T cells and natural killer cells, inhibitangiogenesis, and exert anti-tumor effects (Farber, 1997).

Previous, mainly immunohistochemical, studies in skin biopsies ofMF suggested that both IL-8 (Hansen et al, 1991; McLean Wismeret al, 1994) and IP-10 (Sarris et al, 1995) play an important role in thepreferential migration of neoplastic T cells into the epidermis. Expres-sion of Mig mRNA or protein in CTCL has not been studied thusfar; however, immunostaining by anti-chemokine antibodies on skintissue sections is often difficult to interpret, and does not alwaysreflect the presence of biologically active forms of these chemokines

VOL. 111, NO. 2 AUGUST 1998 CHEMOKINE EXPRESSION IN EPIDERMOTROPIC CTCL 223

(Sticherling et al, 1989). In addition, more recent studies have questionedthe chemotactic properties of IL-8 for T cells (Taub et al, 1993a; Carret al, 1994; Roth et al, 1995).

Because we were interested in whether these chemokines areproduced within the epidermis itself, in this study biopsies frompatients with various stages of MF and large pleomorphic CTCL wereinvestigated for the expression of IL-8 and IP-10, as well as MigmRNA by RNA in situ hybridization (RISH). In addition, weexamined IL-8 receptor A and B (CXCR1 and CXCR2) expressionby immunohistochemistry. The results were correlated with the numberand phenotype of the epidermotropic T cells.

MATERIALS AND METHODS

Patients Skin biopsies were obtained from untreated skin lesions of 17 CTCLpatients, including eight patients with patch/plaque stage MF, four patientswith tumor stage MF, four patients with a large pleomorphic T cell lymphoma,and one patient with a strongly epidermotropic CD81 CTCL. In all cases thediagnosis was based on a combination of clinical, histologic, and immunopheno-typical data, as described previously (Willemze et al, 1997). The neoplastic Tcells had a CD31CD41CD8– T cell phenotype in all cases, except for thesingle patient with a CD81 CTCL. Because IL-8 mRNA expression inpsoriatic epidermis has been well established, skin biopsies from untreatedpsoriatic lesions (n 5 4), uninvolved skin of these psoriatic patients (n 5 4), inaddition to normal human skin of healthy volunteers (n 5 10), were includedas control groups.

Immunohistochemical analysis Five micrometer frozen tissue sectionswere fixed in acetone for 10 min at room temperature, and subsequentlyincubated with appropriately diluted first antibody, biotinylated goat anti-mouseIg antibody (DAKO, Copenhagen, Denmark) and HRP-labeled streptavidin(DAKO). HRP-enzyme activity was revealed using AEC/H2O2 as substratesolution and the sections were counterstained with Mayers hematoxylin andcoverslipped with Aquaperm. The first antibodies used are against CD3 andCD4 (clones BB11 and BF5, respectively, Biosource International, Camarillo,CA), CD8 (clone BC8, Dr. J Wijdenes, Besancon, France), CD54 (cloneMEM111, Monosan, Uden, The Netherlands), CXCR1 (clone 5 A12, Leuko-site, Cambridge, MA), CXCR2 (clone 10H2, Genentech, San Francisco, CA),and Neutrophil Elastase (clone NP57, DAKO).

In situ hybridization

Preparation of 35S labeled RNA probes To obtain IP-10 and Mig encodingcDNA fragment, reverse transcription of oligo(dt) primed total RNA (1 µg)isolated from IFN-γ stimulated primary cultured keratinocytes (Boorsma et al,1994) was performed in 20 µl for 1 h at 42°C using 200 U Superscript IIRT according to the manufacturer’s instructions (Gibco/BRL, Breda, TheNetherlands). Subsequently, two primers (25 pmol each) corresponding to IP-10 mRNA or Mig mRNA (see Table I) were used in a 50 µl polymerasechain reaction (PCR) containing 0.5 U Taq-polymerase (Boehringer Mannheim,Almere, The Netherlands), 0.2 mM of each dNTP and 5 µl first strand cDNA,in 10 mM Tris-HCl (pH 8.3), 50 mM KCl; 1.5 mM MgCl2 [32 cycles of 30@94°C, 19 55°C, and 19 72°C in a thermal cycler (MJ Research, Watertown,MA)]. A single PCR product of the expected size (189 bp for IP-10 and 324bp for Mig) was generated, polished (PCR product polishing kit; Stratagene,La Jolla, CA), and cloned in EcoRV digested pZErO (Invitrogen, Leek, TheNetherlands). Resulting plasmid DNA were sequenced using the dideoxy chaintermination method (Sanger et al, 1977) in order to determine the nature andorientation of the inserts. pZErO plasmid DNA harboring the IP-10 or Migencoding cDNA fragments in either orientation were linearized with BamHIand 35S labeled sense (control) and anti-sense RNA probes were obtained byin vitro transcription with T7 RNA polymerase using the Riboprobe systemaccording to the manufacturer’s instructions (Promega, Leiden, TheNetherlands).

To obtain a cDNA fragment encoding IL-8, 25 pmol of two primerscorresponding to IL-8 mRNA with a RNA polymerase promotor sequenceengineered in the 59 end (T7 for the sense and T3 for the anti-sense primer)were used in a 100 µl PCR using 1 ng of clone 3–10C (Schmid and Weissmann,1987) as a template under the conditions as described above. The PCRgenerated IL-8 encoding cDNA fragment flanked by T3 and T7 RNApolymerase promotors was agarose gel purified in combination with Qiaexspinbind columns (Diagen, Duesseldorf, Germany) and used to synthesize anti-sense (T3) and sense (T7) RNA probes with the Riboprobe system.

After 2 h of transcription, DNA templates were degraded with RQ1 RNase-Free DNase I (10 U; Promega) and the labeled riboprobes were recovered byethanol precipitation using 20 µg yeast t-RNA as carrier, subsequently dissolvedin formamide, and counted. The specific activity of the probes was on averagebetween 5 3 108 and 1 3 109 dpm RNA per µg.

Hybridization procedure Frozen tissue sections (10 µm) were fixed with paraform-aldehyde (4%), acetylated with acetic anhydride in 0.1 M triethanolamine,dehydrated in graded alcohol, treated with chloroform, and air dried. Thereaftersections were overlaid with 100 µl hybridization buffer [50% formamide,43sodium citrate/chloride buffer (SSC), 1 3 Denhardt, 100 mM dithiothreitol,250 µg t-RNA per ml, 100 µg denatured heterologous herring sperm DNAper ml, and 1 µg RNA per ml transcribed from BamHI linearized pZErOwithout insert] containing 1–2 3 104 cpm probe per µl. Sections were mountedwith cover slips and hybridized for 16–18 h at 47–50°C in a humidifiedchamber. After hybridization, nonhybridized probes were removed using highstringency washes with 50% formamide/23SSC at 45°C (twice for 20 min),rinsing in 23SSC, and followed by treatment with RNase A (100 mg per ml)in 23SSC at 37°C (30 min). Next, slides were washed with 50% formamide/23SSC at 45°C (3320 min) and finally two washes with 23SSC at roomtemperature. For autoradiography, dried slides were dipped in Ilford K5 solution,exposed for 1–5 wk, developed using Kodak D19, and counterstained withhematoxylin.

RESULTS

Immunocytochemistry Immunostaining using monoclonal anti-bodies against CD3, CD4, and CD8 was used to assess the degree andnature of the epidermotropic T cells. In skin biopsies from patientswith MF and pleomorphic CTCL, six of 16 cases (nos 1–6) showedmoderate to extensive infiltration of the basal and suprabasal layers ofthe epidermis, five cases (nos 7–11) showed only sparse, often focalinfiltration, whereas in five cases (nos 12–16) epidermotropism wascompletely absent (Table II). The large majority of the epidermotropicT cells were neoplastic CD41 T cells, whereas CD81 T cells weregenerally few or absent. The single patient with a CD81 CTCL (no.17) showed massive epidermal infiltration by neoplastic CD81 T cells,whereas there was only focal infiltration by reactive CD41 T cells.

CD54, which, as IP-10, is expressed by keratinocytes upon activationwith IFN-γ (Dustin et al, 1988), was expressed by epidermal ker-atinocytes in 10 of 17 biopsies. There was a strong correlation betweenCD54 expression and the presence of epidermotropic T cells.

CXCR2 expression was not detected in any of the CTCL specimenstested and CXCR1 expression was only found on dendritic cells inthe epidermis or dermis, and in two cases (nos 2 and 6) on epidermalneutrophils. Thus, the neoplastic T cells expressed neither CXCR1nor CXCR2. In control specimens of lesional psoriatic skin, CXCR2staining was observed on the suprabasal keratinocytes, whereas CXCR1staining was found on dendritic cells and neutrophils. In the epidermisof normal skin of psoriatic patients and healthy controls, staining forCXCR2 was completely negative, whereas CXCR1 was only faintlyexpressed by dendritic cells.

In situ hybridization IP-10 mRNA expression was found in theepidermis, mainly in the basal and suprabasal layers of 11 of 17 biopsies(Table II). IP-10 mRNA expression showed a patchy rather thanuniform pattern (Fig 1). In the group of MF and pleomorphic CTCL,epidermal IP-10 mRNA expression was found in six of six biopsiesshowing moderate to extensive epidermotropism (nos 1–6) and inthree of five biopsies (nos 9–11) showing only focal epidermalinfiltration by neoplastic CD41 T cells, but was not observed in thefive nonepidermotropic CTCL (nos 12–16). Mig mRNA expressionwas found in seven of these 11 epidermotropic MF/pleomorphicCTCL, but not in any of the five nonepidermotropic CTCL. Migexpression co-localized with but was generally less pronounced thanIP-10 expression (Fig 1b). In the CD81 CTCL (no. 17) IP-10 andMig mRNA were expressed only focally, and co-localized withinfiltration of few CD41 reactive T cells, as assessed on serial sections.In the dermal infiltrates variable expression of IP-10 and Mig mRNAwas observed in eight and twelve of 17 CTCL biopsies, respectively.

Focal expression of IL-8 mRNA was found in only two of 15CTCL biopsies tested (nos 2 and 6), and was associated, in both cases,with epidermal necrosis and accumulation of neutrophils.

In contrast, all four psoriatic skin lesions included as positive controlsshowed abundant IL-8 mRNA expression in the upper layers ofthe epidermis, which was generally associated with infiltration ofneutrophils. In normal skin of psoriatic patients and normal controlskin neither IP-10, nor Mig, nor IL-8 mRNA expression were

224 TENSEN ET AL THE JOURNAL OF INVESTIGATIVE DERMATOLOGY

Table I. Sequence and location of PCR primers

Primer Sequence Location on mRNAa

IP-10 sense 59-TGACTCTAAGTGGCATTCAAGG-39 nts 107–128IP-10 anti-sense 59-GATTCAGACATCTCTTCTCACCC-39 nts 275–295Mig sense 59- GTATCTGAGGCACATGTCAGG 39 nts 779–800Mig anti-sense 59- AAAGGCACTGGATTGTGGTAGG 39 nts 1098–1120IL-8 senseb 59AGTGCGTAATACGACTCACTATAGGGTCTGCAGCTCTGTGTGAAGG- 39 nts 136–155IL-8 anti-sensec 59AGTGCGAATTAACCCTCACTAAAGGGCATCTTCACTGATTCTTGG- 39 nts 413–433

aAs described in Luster et al (1985) for IP-10, Farber (1993) for Mig, and Schmid and Weissman (1987) for IL-8.bT7 RNA polymerase promotor sequence is underlined.cT3 RNA polymerase promotor sequence is doubly underlined.

Table II. Relationship between IP-10, Mig, and IL-8 mRNA expression and epidermal infiltration by neoplastic T cells in 17CTCL patients

Epidermotropismb RNA in situ hybridizationc

Diagnosisa Extent CD4 CD8 IP-10 mRNA Mig mRNA IL-8 mRNA

1. MF, patch 111 11 11 11 1 –2. MF, plaque 111 111 occ 11 11 1d

3. MF, plaque 11 11 occ 1 1 –4. MF, plaque 11 11 occ 11 – –5. MF, plaque 11 11 11 11 1 –6. MF, plaque 11 11 occ 1 1 1d

7. MF, plaque 1 1 occ 11 1 ND8. MF, plaque 1 1 – – – –9. MF, tumor 1 1 – 1 1 –10. MF, tumor 1 1 – 1 – –11. MF, tumor 1 1 occ 1 – –12. MF, tumor – – – – – ND13. CTCL, pleomorphic – – – – – –14. CTCL, pleomorphic – – – – – –15. CTCL, pleomorphic – – – – – –16. CTCL, pleomorphic – – – – – –17. CTCL, CD81 111 1F 111 1F 1F –

aMF, mycosis fungoides; CTCL, cutaneous T cell lymphoma; ND, not done; F, focal.bScreening epidermotropism: 111, extensive; 11, moderate; 1, sparse; occ, occasional; –, absent.cScoring RISH: 11, many foci; 1, few foci; –, negative.dAssociated with epidermal necrosis and accumulation of neutrophils.

observed. Sense controls for IP-10 (Fig 1c), Mig, and IL-8 werenegative in all cases.

DISCUSSION

In this study, using RISH, we demonstrated a strong associationbetween epidermal IP-10 mRNA expression and the presence ofintraepidermal T cells. IP-10 mRNA expression was found in 11 of12 epidermotropic CTCL, but not in any of the five nonepidermotropicCTCL. IP-10 mRNA expression correlated with the epidermal infiltra-tion of CD41 neoplastic T cells, rather than reactive CD81 T cells.In accordance, in vitro studies have demonstrated chemoattractiveactivity of IP-10 towards activated CD41 T cells but not towardsCD81 T cells (Taub et al, 1993b). Mig mRNA expression was foundin seven of the 12 epidermotropic CTCL, but not in the any of thefive nonepidermotropic CTCL. In the epidermis, Mig expression co-localized with IP-10 expression, but was generally less pronounced.

The pattern of IP-10, as well as Mig mRNA expression in theepidermis (see also Fig 1), suggests that these chemokines are producedby keratinocytes, although we cannot exclude that intraepidermalleukocytes may also contribute to the chemokine expression. Theproduction of IP-10 mRNA (Boorsma et al, 1994) and protein (Gottliebet al, 1988) by keratinocytes upon stimulation by IFN-γ has been wellestablished. Moreover, it has been shown that keratinocytes producemuch more IP-10 protein compared with endothelial cells, monocytes,and fibroblasts after stimulation with equal amounts of INF-γ (Lusterand Ravetch, 1987).

The distribution of IP-10 and Mig mRNA expression was oftenpatchy, even in cases in which the epidermal CD41 T cells weredistributed more uniformly throughout the epidermis. This phenom-

enon might be due to a characteristic feature of chemokines in thattheir expression is inducible and transient. In recent studies in allergiccontact dermatitis,1 a similar discrepancy between the localization ofIP-10 mRNA (patchy) and the localization of intraepidermal T cellsshowing a more even distribution was observed. In cases withoutepidermal infiltration by CD31CD41 T cells, however, epidermal IP-10 mRNA expression was never found. The association betweenepidermal IP-10 mRNA expression and the presence of CD41 T cellsboth in CTCL and in allergic contact dermatitis suggests that IP-10plays an important role in the preferential infiltration of both benignand neoplastic CD41 cells in the epidermis.

This conclusion is consistent with the results of immunohistochemicalstudies reported by Sarris et al (1995). Using a polyclonal antibodyagainst IP-10, these authors noted increased expression of IP-10protein in the epidermis of epidermotropic CTCL; however, innonepidermotropic CTCL as well as in normal human skin, basal layerkeratinocytes consistently expressed IP-10 protein. In contrast, in ourstudy IP-10 mRNA was not observed in the epidermis of normal humanskin, or in nonepidermotropic CTCL. The constitutive expression ofIP-10 protein in the basal layers of normal epidermis, as found bySarris et al (1995), is unexplained.

Concerning Mig, our results suggest that epidermal expression ofthis chemokine might also contribute to epidermal infiltration byCD41 cells. Because previous in vitro and in vivo studies showed thatIFN-γ is the only inducer of Mig (Farber, 1997), our observations

1Flier J, van der Stoop PM, van Beek PJ: IP-10 mRNA is expressed in allergicbut not in irritant contact dermatitis. J Invest Dermatol 109:451, 1997 (abstr.)

VOL. 111, NO. 2 AUGUST 1998 CHEMOKINE EXPRESSION IN EPIDERMOTROPIC CTCL 225

Figure 1. Abundant epidermal IP-10 and Mig mRNA expression inearly MF lesion. In situ hybridization with 35S-UTP-labeled IP-10 anti-senseRNA (a), Mig anti-sense (b), and IP-10 sense RNA probe (c) followed byemulsion autoradiography was performed on serial cryostat sections as describedin Materials and Methods. (a, b) Large foci of silver grains representing sites ofIP-10 and Mig mRNA transcripts in the epidermis. (c) Absence of silver grainsin the epidermis with sense probe showing the specificity of the in situhybridization procedure. Scale bar, 45 µm.

further indicate that INF-γ is an important inducer of chemokineexpression in epidermotropic CTCL.

Previous studies suggested that IL-8 is responsible for the infiltrationof T cells into the epidermis of CTCL (Hansen et al, 1991; McLeanWismer et al, 1994). Using a polyclonal antibody against IL-8, Hansenet al demonstrated elevated IL-8 immunoreactivity in the epidermis ofsix CTCL patients, and suggested that IL-8, by its chemoattractiveproperties, is responsible for the migration of T cells towards theepidermis. More recently, McLean Wismer et al (1994), using the

monoclonal antibody 52E8, raised against IL-8 (Sticherling et al, 1989)as well as RISH, reported intense epidermal expression of IL-8 mRNAand protein in 14 of 15 CTCL lesions, compared with negligiblereactivity in normal human epidermis; however, a correlation betweenIL-8 expression and epidermotropism was not made.

In contrast, in this study focal IL-8 mRNA expression was detectedin the epidermis of only two of 15 CTCL biopsies, which is consistentwith the observations of Volc-Platzer et al who could not detect IL-8mRNA in a group of five epidermotropic T cell lymphoma.2 Theseand other observations strongly argue against a role of IL-8 in epidermalT cell infiltration in CTCL. First, in both CTCL cases demonstratingepidermal IL-8 mRNA expression, this expression was associatedwith epidermal necrosis and accumulation of CXCR1 expressingneutrophils. Second, the reliability of our RISH procedures wasconfirmed by demonstration of IL-8 mRNA in the upper layers ofthe epidermis in four untreated psoriasis lesions. An identical expressionpattern has been reported previously (Gillitzer et al, 1991; Kulke et al,1996). Third, using the monoclonal antibody 52E8 raised against IL-8(kindly provided by Dr. M. Sticherling, Department of Dermatology,University of Kiel, Kiel, Germany), which was also used in the studyof McLean Wismer et al (1994), we found no or minimal epidermalstaining in the epidermis of the epidermotropic CTCL. In contrast, auniform suprabasal epidermal staining was found in all five nonepiderm-otropic CTCL as well as in normal skin (data not shown). Identicalresults in normal skin have been reported by Sticherling et al (1991),which have resulted in the view that antibody 52E8 may react with astorage form of IL-8, but not with biologically active IL-8. Fourth, ifIL-8 plays a role in epidermotropism, one would expect that theepidermotropic T cells express receptors for IL-8. Using monoclonalantibodies against CXCR1 and CXCR2, however, intraepidermal anddermal CD41 T cells were consistently negative for both receptorantibodies. Previous studies using these antibodies against CXCR1 andCXCR2, demonstrated expression of CXCR2 in upper keratinocytesin lesional psoriatic skin (Beljaards et al, 1997; Bornscheuer et al, 1997).The observation that the neoplastic CD41 T cells in these CTCL donot stain with antibodies against CXCR1 and CXCR2 is consistentwith recent studies demonstrating that CXCR1 and CXCR2 areexpressed by only a small proportion of CD81 T cells, but not byCD41 T cells (Chuntarapai et al, 1994; Qin et al, 1996), and underscorerecent doubt as to the chemoattractive activity of IL-8 towards CD41

T cells (Taub et al, 1993a; Carr et al, 1994; Roth et al, 1995).In conclusion, the results of this study suggest that IP-10, and to a

lesser extent Mig, but not IL-8 is involved in the preferential infiltrationof neoplastic CD41 T cells in CTCL.

REFERENCES

Baggiolini M, DeWald B, Moser B: Interleukin-8 and related chemotactic cytokines-CXC and CC chemokines. Adv Immunol 55:97–179, 1994

Bazan JF, Bacon KB, Hardiman G, et al: A new class of membrane-bound chemokinewith a CX3C motif. Nature 385:640–644, 1997

Beljaards RC, Meijer CJLM, van der Putte SCJ, Hollema H, Geerts ML, Bezemer PD,Willemze R: Primary cutaneous T cell lymphoma: clinicopathologhic features andprognostic parameters of 35 cases other than mycosis fungoides and CD3-positivelarge cell lymphoma. J Pathol 172:53–60, 1994

Beljaards RC, van Beek P, Nieboer C, Stoof TJ, Boorsma DM: The expression of theinterleukin-8 receptor in untreated and treated psoriasis. Arch Dermatol Res 289:440–443, 1997

Boorsma DM, de Haan P, Willemze R, Stoof TJ: Human growth factor (huGRO),interleukin-8 (IL-8) and interferon-γ-inducible protein (γ-IP-10) gene expression incultured normal human keratinocytes. Arch Dermatol Res 286:471–475, 1994

Bornscheuer E, Schroder J-M, Christophers E, Sticherling M: Psoriatic keratinocytesexpress the interleukin 8 receptor B. Arch Dermatol Res 289S:A57, 1997

Carr MW, Roth SJ, Luther E, Rose SS, Springer TA: Monocyte chemoattractant protein1 acts as a T-lymphocyte chemoattractant. Proc Natl Acad Sci USA 91:3652–3656, 1994

Chuntarapai A, Lee J, Hebert CA, Kim KJ: Monoclonal antibodies detect differentdistribution patterns of IL-8 receptor A and IL-8 receptor B on human peripheralblood leukocytes. J Immunol 153:5682–5688, 1994

Dustin ML, Singer KH, Tuck DT, Springer TA: Adhesion of T lymphocytes to epidermalkeratinocytes is regulated by interferon γ and is mediated by intercellular adhesionmolecule 1 (ICAM-1). J Exp Med 167:1323–1340, 1988

2Volc-Platzer B, Berger R, Mielke V, Wolff K, Stingl G, Gillitzer R: NAP-1/IL-8 is chemotactic for neutrophils but not for T-lymphocytes in vivo. ArchDermatol Res 284:27, 1992 (abstr.)

226 TENSEN ET AL THE JOURNAL OF INVESTIGATIVE DERMATOLOGY

Edelson RL: Cutaneous T cell lymphoma: mycosis fungoides, Sezary syndrome, and othervariants. J Am Acad Dermatol 2:89–106, 1980

Farber JM: HuMig. a new human member of the chemokine family of cytokines. BiochemBiophys Res Commun 192:223–230, 1993

Farber JM: Mig and IP-10: CXC chemokines that target lymphocytes. J Leukocyte Biol61:246–257, 1997

Gillitzer R, Berger R, Mielke V, Muller C, Wolff K, Stingl G: Upper keratinocytes ofpsoriatic skin lesions express high levels of NAP-1/IL-8 mRNA in situ. J InvestDermatol 97:73–79, 1991

Gottlieb AB, Luster AD, Posnett DN, Carter DM: Detection of a γ interferon-inducedprotein IP-10 in psoriatic plaques. J Exp Med 168:941–948, 1988

Hansen ER, Vejlsgaard GL, Lisby S, Heidenheim M, Baadsgaard O: Epidermal interleukin1α functional activity and interleukin 8 immunoreactivity are increased in patientswith cutaneous T cell lymphoma. J Invest Dermatol 97:818–823, 1991

Kulke R, Todt-Pingel I, Rademacher D, Rowert J, Schroder J-M: Co-localizedoverexpression of GRO-α and IL-8 mRNA is restricted to the suprapapillary layersof psoriatic lesions. J Invest Dermatol 106:526–530, 1996

Loetscher M, Gerber B, Loetscher P, et al: Chemokine receptor specific for IP10 and Mig:structure, function, and expression in activated T-lymphocytes. J Exp Med 184:963–969, 1996

Luster AD, Ravetch JV: Biochemical characterization of a γ interferon-inducible cytokine(IP-10). J Exp Med 166:1084–1097, 1987

Luster AD, Unkeless JC, Ravetch JV: γ-Interferon transcriptionally regulates an early-response gene containing homology to platelet proteins. Nature 315:672–676, 1985

McLean Wismer J, McKenzie RC, Sauder DN: Interleukin-8 immunoreactivity inepidermis of cutaneous T cell lymphoma patients. Lymphokine Cytokine Res 13:21–27, 1994

Miller MD, Krangel MS: Biology and biochemistry of the chemokines: a family ofchemotactic and inflammatory cytokines. Crit Rev Immunol 12:17–46, 1992

Murphy PM: The molecular biology of leukocyte chemoattractant receptors. Ann RevImmunol 12:593–633, 1994

Oppenheim JJ, Zachariae COC, Mukaida N, Matsushima K: Properties of the novel

proinflammatory supergene ‘‘intercrine’’ cytokine family. Ann Rev Immunol 9:617–648, 1991

Qin S, LaRosa G, Campbell JJ, et al: Expression of monocyte chemoattractant protein-1and interleukin-8 receptors on subsets of T cells: correlation with transendothelialchemotactic potential. Eur J Immunol 26:640–647, 1996

Roth SJ, Carr MW, Springer TA: C-C chemokines, but not the C-X-C chemokinesinterleukin-8 and interferon-gamma inducible protein-10, stimulate trans-endothelialchemotaxis of T lymphocytes. Eur J Immunol 25:3482–3488, 1995

Sanger F, Nicklen S, Coulson AR: DNA sequencing with chain-terminating inhibitors.Proc Natl Acad Sci USA 74:5463–5467, 1977

Sarris AH, Esgleyes-Ribot T, Crow M, et al: Cytokine loops involving interferon-gammaand IP-10, a cytokine chemotactic for CD41 lymphocytes: an explanation for theepidermotropism of cutaneous T cell lymphoma. Blood 86:651–658, 1995

Schmid J, Weissmann C: Induction of mRNA for a serine protease and a β-thromboglobulin-like protein in mitogen-stimulated human leukocytes. J Immunol139:250–256, 1987

Sticherling M, Schroder JM, Christophers E: Production and characterization of monoclonalantibodies against the novel neutrophil activating peptide NAP/IL-8. J Immunol143:1628–1634, 1989

Sticherling M, Bornscheuer E, Schroder JM, Christophers E: Localization of neutrophil-activating peptide-1/interleukin-8-immunoreactivity in normal and psoriatic skin.J Invest Dermatol 96:26–30, 1991

Taub DD, Conlon K, Lloyd AR, Oppenheim JJ, Kelvin DJ: Preferential migration ofactivated CD41 and CD81 T cells in response to MIP-1 alpha and MIP-1 beta.Science 260:355–358, 1993a

Taub DD, Lloyd AR, Conlon K, et al: Recombinant human interferon-inducible protein10 is a chemoattractant for human monocytes and T lymphocytes and promotes Tcell adhesion to endothelial cells. J Exp Med 177:1809–1814, 1993b

Wells TNC, Peitsch MC: The chemokine information source: identification andcharacterization of novel chemokines using the WorldWideWeb and expressedsequence tag databases. J Leukocyte Biol 61:545–550, 1997

Willemze R, Kerl H, Sterry W, et al: EORTC classification for primary cutaneouslymphomas: a proposal from the Cutaneous Lymphoma Study Group of the EuropeanOrganization for Research and Treatment of Cancer. Blood 90:354–360, 1997