Upregulation of IL-6, IL-8, CXCL1, and CXCL2 dominates gene expression in human fibroblast cells exposed to Loxosceles reclusa sphingomyelinase D: insights into spider venom dermonecrosis.

TO THE EDITOR Brown recluse (Loxosceles spp.) spiders are arachnid species known to cause necrotic arachnidism. The envenomation, described as loxoscelism, is associated with localized pain, erythema, and edema followed by the development of necrosis (Futrell, 1992; Hogan et al., 2004). However, the specific pathophysiological mechanisms by which Loxosceles venom exerts these noxious symptoms are multifactorial and not fully understood. The causative factor for production of necrotic lesions is generally considered to be the enzyme sphingomyelinase D (SMD), which cleaves sphingomyelin to form choline and ceramide 1-phosphate. Four active forms of SMD with molecular mass of 32,000 are found in L. recluse (Kurpiewski et al., 1981; Futrell, 1992). Sphingomyelinases of comparable size are also described in the venoms of L. intermedia, L. gaucho, and L. laeta (Silva et al., 2004). Pedrosa et al. (2002) reported that initial incubation with specific antiserum of biologically functional recombinant L. laeta SMD inhibited dermonecrotic activity. The fact that SMD enzymatic activity is necessary for initiation of the necrosis was confirmed by Ramos-Cerrillo et al. (2004). They analyzed three recombinant L. boneti SMD isoforms for enzymatic activity in vitro, and dermonecrosis when injected in rabbits. Isoforms 1 and 2 retained enzymatic activity and caused dermonecrosis in vivo. Interestingly, isoform 3 which did not produce dermonecrosis was also enzymatically inactive. Similar results for SMD isoforms from L. intermedia were previously reported by Tambourgi et al. (1998). It is currently thought that SMD does not directly induce necrosis but rather it acts as initiator of the pathology by stimulating inflammatory response in endothelial cells (Patel et al., 1994; Desai et al., 1999, 2000). We hypothesize that the stromal fibroblasts are also involved in loxosceles pathophysiology. To test this as well as further investigate in vitro the mechanisms associated with loxoscelism, we assessed the gene expression profiles of SMD-treated and untreated human fibroblasts using GeneChips, protein microarrays, and quantitative reverse transcription-PCR. Human fibroblasts cells (ATCC CRL-1635) maintained in DMEM, containing 10% serum were used in these experiments. At the time of treatment, the media was replaced with serum-free DMEM. After 4 hours of incubation the media was changed and the cells were cultured with 1 g/ml L. recluse isoform I recombinant SMD (Olvera et al., 2006) in DMEM without serum for 30 minutes, 1, 3, or 5 hours. At this concentration of toxin, there was no decernable cell injury, and greater than 95% of cells remained viable (by trypan blue exclusion assay) after 48 hours of incubation. This concentration for the SDM toxin was chosen based on a previously report (Gomez et al., 2001) where an intradermal application of 3 g complete L. recluse venom showed reproducible pathology in the rabbit model of envenomation. Using AMPLEX assay (Invitrogen, Carlsbad, CA), we found that of 1 g of SMD showed sphingomyelinase activity corresponding to the enzyme activity of 3 g of our venom preparation. All experiments were approved by the Institutional Biosafety Committee of the University of Virginia. The gene expression profiles of two samples after 5 hours of SMD treatment were compared with two controls using Affymetrix Hg-U133A arrays. To identify the differentially expressed genes between treated and the control samples chip data were analyzed using the Affymetrix Data Mining Tool. Similar experiments were performed using whole L. recluse venom (5 g/ml) and comparable changes in expression were observed as to those in with the SDM toxin (data not shown) suggesting that the SDM toxin is the primary contributor to changes in fibroblast gene expression. Three groups of upregulated genes from the SDM-treated fibroblast gene expression experiments were considered of further interest in relation to the molecular mechanisms associated with loxosceles pathology: human cytokines (IL-6, IL-8, IL-1B, CXCL1, CXCL2, CCL5, tumor necrosis factor); genes involved in glycosphingolipid metabolism pathway (alpha 1,4-galactosyltransferase, UDP-glucose ceramide glucosyltransferase); and proteins known as factors of transcriptional regulation (nuclear factor of kappa light polypeptide gene enhancer in B-cells 2 (p49/p100), nuclear factor of kappa light polypeptide gene enhancer in B-cells inhibitor, epsilon). The expression profiles of seven arbitrarily selected regulated cytokines and NF-B were validated by qPCR (Table 1). To determine if the alterations in transcription actually lead to increase protein expression media from treated and untreated fibroblasts (5 hours) were analyzed with FASTQuant Human II protein microarrays (Table 1). To extend the analysis of the upregulated cytokines, expression levels at four time points following treatment were investigated. Changes were first observed at 3 hours following SMD treatment, and at 5 hours significant upregulation of IL-6, IL-8, CXCL-1, and CXCL-2 was observed (Figure 1). In summary our study focused on the global gene expression changes in human fibroblast cells, following incubation with recombinant SMD. We detected induction of a strong inflammatory response involving transcriptional upregulation of IL-6, IL-8, IL-1B, CCL5, CXCL1, CXCL2, and tumor necrosis factor- in conjunction with a proportional increase in the expression of some of these proteins. We also observed a modest increase in the transcription of two genes involved in glycosphingolipid metabolism pathway (UDP-glucose ceramide glucosyltransferase and -1,4-galactosyltransferase), as well as NF-B and NFKBIA (NF-B inhibitor). In contrast to a previous study conducted on endothelial cells (Patel et al., 1994), we found IL-6 significantly upregulated in fibroblast cultures treated with SMD. IL-6 and IL-8 expression is known to be activated via the transcriptional factor NF-B (Matsusaka et al., 1993). It is also noted in the literature (Bauerle and Henkel, 1994; Ballou et al., 1996) that phosphorylated ceramide is an important factor in the activation of the transcriptional factor NF-B. In the normal acute inflammatory process, neutrophils are the first cells recruited by the immune system, followed by their clearance and mononuclear cells infiltration. IL-6 is one of the regulators of this immunological switch (Kaplanski et al., 2003; Jones, 2005). Thus, based on the results presented and the literature we speculate that SMD hydrolyzes sphingomyelins localized in the outer leaflet of the plasma membrane of fibroblasts and ceramide phosphate, which is the product of sphingomyelin hydrolysis, is responsible for an abnormal activation of NF-B. Furthermore, we suggest that secretion of ceramide phosphate-induced IL-8 by endothelial cells and stromal fibroblasts in addition to the normally immunoregulated IL-8, might be connected to loxoscelism histopathology dominated, by dense neutrophilic infiltrate, necrosis, edema, and proliferation of fibroblasts (Elston et al., 2000; Tambourgi et al., 2005). Therefore, further in vivo investigations focused on the control of IL-8 expression after envenomation of spider venom or SMD could be of critical to the understanding of the pathophysiology of loxoscelism and indentify novel mechanism for therapeutic intervention. The authors state no conflict of interest.