To the Editor:

We appreciate Dr. Sfriso and colleagues' comments on our work. In our study we demonstrated by immunohistochemistry that IP-10 and Mig are expressed in the ductal epithelial cells adjacent to lymphoid infiltrates of the salivary gland in patients with SS, and we showed by reverse transcriptase–polymerase chain reaction (RT-PCR) and enzyme-linked immunosorbent assay (ELISA) that cultured SS salivary epithelial cells produce messenger RNA (mRNA) and proteins of IP-10 and Mig in response to interferon-γ (IFNγ). Moreover, most infiltrating CD3+ lymphocytes in dense periductal foci express CXCR3, the receptor for IP-10 and Mig. Our findings suggest that IFNγ and IFNγ-induced T cell–attracting chemokines are involved in the development of lymphocytic infiltrates in the SS salivary gland. In contrast to our results, Sfriso et al found that 1) CXCR3+ lymphocytes existed not only in SS salivary glands but also in the salivary glands of control subjects, 2) IP-10 was present in the ductal epithelial cells of the majority of control subjects, and 3) CXCR3 was expressed in SS and normal ductal epithelial cells. Their work was performed using immunohistochemistry.

As pointed out by Sfriso et al, the discrepancy regarding the finding of CXCR3+ lymphocytes in the normal salivary gland could be explained by the difference in the selection of control subjects between our study and theirs. Control subjects in our study reported sicca symptoms, but were completely negative for any immunologic abnormalities, such as autoantibodies including anti-SSA or anti-SSB, antinuclear antibodies, or rheumatoid factor, and hypergammaglobulinemia. The focus score in all control subjects was 0. Another possibility is that a few lymphocytes may be present in the normal salivary gland due to tissue injury occurring during the biopsy procedure. In our study we showed, by flow cytometry, that peripheral blood lymphocytes (PBLs), even in a quiescent state, express CXCR3, though the CXCR3 expressed on freshly isolated PBLs is not functionally active. Thus, it is not surprising that a few CXCR3+ lymphocytes could be found in biopsy specimens from some control subjects.

Sfriso et al found IP-10 positivity in normal ductal epithelium as well as in SS ductal epithelium. Although their finding is interesting, we believe it needs to be confirmed. Immunohistochemistry is a powerful tool for the detection of proteins in the tissue; however, false-positive results are always possible. In order to confirm their finding, other techniques besides immunohistochemistry should be used, for example, RT-PCR combined with tissue microdissection (1). We used cultured SS salivary epithelial cells from minor salivary glands and confirmed the production of IP-10 and Mig proteins by ELISA in addition to immunohistochemistry. The production of IP-10 and Mig from normal salivary epithelial cells is completely dependent on exogenous IFNγ, and IFNγ was shown to be absent from salivary tissue of healthy subjects (1). If IP-10 is produced by the ductal epithelial cells of control subjects, the substance(s) that stimulate this production needs to be identified.

Perhaps the most notable and controversial finding in the study by Sfriso et al is the expression of CXCR3 in SS and normal ductal epithelial cells. It is well known that functional chemokine receptors are expressed in some types of epithelial cells. For example, CXCR4, the receptor for stromal cell–derived factor 1 (SDF-1), has been shown to be expressed in human alveolar epithelial cells (2) and in human colonic epithelial cells (3). CXCR4 is involved in organogenesis as well as in cell migration, as evidenced by the fact that CXCR4−/− mice die in utero and are defective in cerebellar and vascular development, hematopoiesis, and cardiogenesis, as are SDF-1−/− mice (4, 5). In contrast, with rare exceptions, such as neurons (6), functional CXCR3 expression is limited to leukocytes. Very recently, CXCR3 was shown to be expressed on the apical side of normal human renal proximal tubules, by RT-PCR combined with microdissection and immunohistochemistry (7). Thus, it is possible that human salivary ductal epithelial cells express functional CXCR3.

To examine CXCR3 expression in the ductal epithelium, we performed RT-PCR analysis for CXCR3 mRNA in salivary epithelial cells from 2 SS patients and a healthy control subject, in the presence or absence of IFNγ as described in our report. Total RNA was extracted from the cells, reverse-transcribed, and analyzed by PCR for expression of CXCR3 mRNA. Amplification was carried out for 30 cycles, which included 1 minute at 92°C, 1 minute at 55°C, and 1 minute at 72°C. Amplified DNA was electrophoresed on a 2% agarose gel, stained with ethidium bromide, visualized under ultraviolet light, and photographed. GAPDH was used as a housekeeping gene. Primer sequences were as follows: for GAPDH, sense 5′-TCC-ATG-ACA-ACT-TTG-GTA-TCG-3′ and antisense 5′-GTC-GCT-GTT-GAA-GTC-AGA-GGA-3′; for CXCR3, sense 5′-AGA-GGA-CGC-TGT-CTT-TGC-AT-3′ and antisense 5′-GTC-CTT-TCA-CCC-ACC-TTT-CA-3′. Expected sizes of the PCR products were 376 bp for GAPDH and 383 bp for CXCR3. As shown in Figure 1, with IFNγ stimulation, weak expression of CXCR3 mRNA was demonstrated in specimens from the 2 SS patients. Because there is no lymphocyte contamination in the salivary epithelial cell population, CXCR3 mRNA could thus be considered to be expressed by epithelial cells. This finding seems to support the results described by Sfriso et al. However, mRNA expression does not always correlate with protein production. Unlike CXCR4−/− mice, mice lacking CXCR3 seem to be phenotypically and developmentally normal (8).

thumbnail image

Figure 1. Reverse transcriptase–polymerase chain reaction determination of the expression of CXCR3 mRNA, with or without treatment for 12 hours with 1,000 units/ml interferon-γ (IFNγ), in the salivary epithelial glands of 2 patients with Sjögren's syndrome (SS) and 1 normal control subject.

Download figure to PowerPoint

CXCR3 expression in the salivary ductal epithelial cells needs to be confirmed by other methods. Furthermore, it would be necessary to ascertain whether the receptor is functional, for example, by demonstrating intracellular calcium mobilization in salivary epithelial cells with CXCR3 ligand stimulation.


  1. Top of page
  • 1
    Sun D, Emmert-Buck MR, Fox PC. Differential cytokine mRNA expression in human labial minor salivary glands in primary Sjögren's syndrome. Autoimmunity 1997; 28: 12537.
  • 2
    Murdoch C, Monk PN, Finn A. Functional expression of chemokine receptor CXCR4 on human epithelial cells. Immunology 1999; 98: 3641.
  • 3
    Jordan NJ, Kolios G, Abbot SE, Sinai MA, Thompson DA, Petraki K, et al. Expression of functional CXCR4 chemokine receptors on human colonic epithelial cells. J Clin Invest 1999; 104: 10619.
  • 4
    Zou YR, Kottmann AH, Kuroda M, Taniuchi I, Littman DR. Function of the chemokine receptor CXCR4 in haematopoiesis and cerebellar development. Nature 1998; 393: 5959.
  • 5
    Tachibana K, Hirota S, Iizawa H, Yoshida H, Kawabata K, Kataoka Y, et al. The chemokine receptor CXCR4 is essential for vascularization of the gastrointestinal tract. Nature 1998; 393: 5914.
  • 6
    Xia MQ, Bacskai BJ, Knowles RB, Qin SX, Hyman BT. Expression of the chemokine receptor CXCR3 on neurons and the elevated expression of its ligand IP-10 in reactive astrocytes: in vitro ERK 1/2 activation and role in Alzheimer's disease. J Neuroimmunol 2000; 108: 22735.
  • 7
    Bek MJ, Reinhardt HC, Fischer KG, Hirsch JR, Hupfer C, Dayal E, et al. Up-regulation of early growth response gene-1 via the CXCR3 receptor induces reactive oxygen species and inhibits Na+/K+-ATPase activity in an immortalized human proximal tubule cell line. J Immunol 2003; 170: 93140.
  • 8
    Hancock WW, Lu B, Gao W, Csizmadia V, Faia K, King JA, et al. Requirement of the chemokine receptor CXCR3 for acute allograft rejection. J Exp Med 2000; 192: 15159.

Noriyoshi Ogawa MD, PhD*, Li Ping MD*, Susumu Sugai MD, PhD*, * Kanazawa Medical University, Ishikawa, Japan.