Detection of autoantibodies to DSF70/LEDGFp75 in Mexican Hispanics using multiple complementary assay platforms

The presence of antinuclear autoantibodies (ANA) is a key feature of ANA-associated rheumatic diseases (AARD) such as systemic lupus erythematosus and scleroderma [1]. The dense fine speckled (DFS) ANA pattern has recently been the subject of intense investigation since it is one the most commonly recognized autoantibody patterns produced by human sera referred to clinical laboratories for ANA testing by indirect immunofluorescence (IIF) microscopy in HEp-2 substrates [2‐5]. This pattern is characterized by uniformly distributed DFS in interphasic nuclei, staining of mitotic chromosomes, and reactivity against a 70–75 kDa protein by immunoblotting [6].

Anti-DFS autoantibodies specifically target the nuclear protein DFS70, most commonly known as lens epithelium-derived growth factor p75 (LEDGF/p75) [7‐9]. DFS70/LEDGFp75 is a stress response transcription coactivator that protects mammalian cells against diverse environmental stressors [8]. This autoantigen is also essential for the integration of human immunodeficiency virus 1 (HIV-1) [10]. In addition, it has been recognized as an oncoprotein whose overexpression in cancer cells promotes tumor aggressive properties such as increased clonogenicity, migration, invasion, chemotherapy resistance, stress survival, angiogenesis and tumor growth (reviewed in Refs. [8, 9]).

The clinical and biological significance of autoantibodies to DFS70/LEDGFp75 remains enigmatic, although it is plausible that these antibodies arise in response to molecular and cellular events associated with altered structure or expression of this antigen under a pro-inflammatory microenvironment [8]. While these autoantibodies have been reported at relatively low frequencies (<5%) in patients with AARD and inflammatory myopathies (IIM), their prevalence in apparently healthy individuals (HI) and young women appears to be higher [2‐4, 8, 11‐14]. They have also been detected at varied frequencies in non-rheumatic conditions such as atopic dermatitis, interstitial cystitis, prostate cancer, and eye diseases [3, 4, 6, 8]. When present in patients with AARD, anti-DFS70/LEDGFp75 antibodies are typically accompanied by other ANA, but in HI and patients with non-AARD conditions they tend to be monospecific, hence becoming attractive clinical biomarkers to potentially exclude a diagnosis of AARD [12‐17].

Recent studies on anti-DSF70/LEDGFp75 autoantibodies have highlighted inter-laboratory discrepancies in their detection and frequencies, likely due to differences in assay platforms and human expertise in the interpretation of the IIF-DFS ANA pattern, or, possibly, variations in the gender, ethnicity, geographical location, and environmental exposures of the populations screened for the presence of these autoantibodies [4, 8]. While their frequency has been widely examined in populations of European and Asian descent, as well as in Brazilians, their frequency in Hispanic/Latino populations has not been reported. In this study, we examined for the first time the frequency of anti-DSF70/LEDGFp75 autoantibodies in sera from Mexican patients with RA, dermatomyositis (DM), and obesity (OB), as well as in Mexican HI. Multiple complementary assay platforms were used in order to develop consensus regarding the types of techniques/platforms that provide accurate detection and confirmation of these autoantibodies.

The presence of anti-DFS70/LEDGFp75 autoantibodies in apparently HI and in a broad spectrum of non-AARD inflammatory conditions have made difficult to pinpoint their clinical and biological significance [4, 5, 8]. Thus far, the strongest associations found for these antibodies are younger age and female gender (reviewed in Ref. [4]), but these results need to be confirmed with very large cohorts from different geographic locations and standardized antibody detection methods. There is also evidence suggesting that the relatively low frequency (<5%) of these antibodies, particularly when they are the sole ANA pattern in serum, in AARD makes them potentially useful biomarkers to rule out the presence of systemic autoimmune disease, although they may not help in ruling out RA [4, 14‐17].

It is still unclear what the presence of these autoantibodies is telling us [4, 8]. Our group proposed recently that depending on the context in which they arise these antibodies could play protective or pathogenic roles, or serve as “sensors” or “reporters” of increased oxidative stress or inflammatory cellular damage associated with DFS70/LEDGFp75 upregulation or proteolytic cleavage [8]. Given the enigmatic nature of anti-DFS70/LEDGFp75 autoantibodies, it was also emphasized that when assessing their frequency in different populations, investigators and clinicians should carefully consider the individual’s health history, ethnicity, geographic location, lifestyle, and exposure to environmental stressors [8].

While the frequency and properties of anti-DFS70/LEDGFp75 autoantibodies have been widely investigated in European, Caucasian American, European, Brazilian, and Asian populations [4, 8], to our knowledge their frequency has not been determined in relatively homogeneous Hispanic/Latino populations with and without AARD. The present study was designed to determine whether the reported low frequencies of these autoantibodies in rheumatic diseases and relatively elevated frequencies in HI can be also reproduced in a Hispanic Mexican cohort that included healthy individuals as well as patients with obesity, DM, and RA. To ensure the accurate detection of these autoantibodies, we initially tested for their presence using the HEp-2 based ANA-IIF test and then confirmed the results using three different but complementary detection platforms. These confirmatory tests were necessary in light of growing concerns that the recognition of the DFS pattern by the ANA-IIF test still remains challenging in the clinical laboratory setting [7, 21‐25].

Recently, it was emphasized at the 2nd ICAP Workshop and the 2nd International Autoantibody Standardization (IAS) Workshop, both conducted in conjunction with the 12th Dresden Symposium on Autoantibodies, that the DFS-IIF pattern, defined as the “AC-2 pattern” by the ICAP, should be carefully differentiated from the homogenous as well as other speckled nuclear patterns, including the NFS pattern [26]. Conrad and colleagues [26] pointed that the identification of the DFS pattern by HEp-2 IIF is not sufficient for accurate detection of anti-DFS70/LEDGFp75 autoantibodies, and confirmatory assays such as CIA and ELISA, or HEp-2 IIF with serum pre-adsorption using recombinant DFS70/LEDGFp75 protein, are required. Routine confirmation of the presence of these autoantibodies in the clinical setting is likely to prevent additional unnecessary tests for AARD diagnosis, resulting in cost-effective patient management. For instance, a recent study by Gundín et al. [27] conducted in Spain implemented a new workup ANA algorithm that included testing for the presence of anti-DFS70/LEDGFp75 antibodies, using confirmatory assays, to clinically discriminate AARD from non-AARD patients in ANA-IIF positive individuals, resulting in significant cost-effective patient management in their setting.

In our initial HEp-2 IIF analysis, we found several sera that produced a nuclear staining pattern consistent with the description of the NFS pattern. It was therefore necessary to distinguish these NFS sera from the DFS sera by conducting confirmatory tests to validate the presence of autoantibodies to DFS70/LEDGFp75 in the DFS sera and their absence in the NFS sera. We found a strong concordance in the detection of anti-DFS70/LEDGFp75 by the different confirmatory tests in the DFS sera, highlighting the growing consensus that confirmation of a positive DFS pattern by the ANA IIF-HEp-2 test should be adequately achieved with additional tests [4, 5, 7, 21, 22, 24, 25]. However, we suggest that confirmatory WB data should be interpreted with caution given that whole lysates prepared from different cell lines may have varying expression levels of the DFS70/LEDGFp75 protein [19, 20], which could lead to inter-laboratory variations in DFS serum reactivity if the same cell line is not universally used. In our experience, cancer cell lines such as Jurkat, PC3, DU145, HeLa (including its derivative HEp-2), HCT116, and U2OS express elevated levels of DFS70/LEDGFp75, compared to non-cancer or non-transformed cell lines, thus providing excellent sources of this antigen for recognition by high titer anti-DFS antibodies [19, 20]. A challenge that we encountered in our study, however, was that low titer anti-DFS sera reacted very weakly against a 70–75 kDa protein band by WB in both Jurkat and PC3 cells, and often with high background. An alternative to using whole cell lysates for WB detection of anti-DFS antibodies, is to confirm the DFS-IIF pattern by dot blot or line blot methods that use the full length DFS70/LEDGFp75 or a truncated C-terminal fragment of this protein containing the autoepitope region [22].

In our study, the DFS70-ELISA (MBL) and DFS70-CIA (Inova) platforms showed excellent detection sensitivity and concordance between the anti-DFS values obtained, in spite of the differences in the DFS70/LEDGFp75 antigens used, which is consistent with previous results [14]. Some studies have reported a high frequency (>15%) of anti-DFS70/LEDGFp75 antibodies detected by ELISA (non-MBL) in healthy controls as well as in various non-AARD disease conditions [6, 28‐30]. It should be noted that in many cases these high frequencies have not been confirmed independently using DFS70-CIA or ELISA-MBL, suggesting that the source and form of the DFS70/LEDGFp75 antigen used, type of equipment or assay used for detection, and the level of stringency used to calculate the cut-off values may influence these frequencies.

The relatively low frequency of anti-DFS70/LEDGFp75 autoantibodies in the disease groups that we tested by the different assay platforms (1.4% in DM, 4.3% in RA) is consistent with previous frequencies reported in the literature for rheumatic diseases [4, 8]. The 6.7% frequency found in OB patients is higher than that typically found in patients with AARD (<5%), but close to the average for HI (6.45% average from 14 studies, range 0–21.6%) [4]. Consistent with previous studies in other ethnic populations, we observed the highest frequency in apparently HI (17.4%); however, this value is close to the top of the 0–21.6% range reported in the literature [4]. These results suggest that geographic differences or ethnicity may not influence the frequency of these autoantibodies. However, given that our sample size for apparently HI (n = 23) was relatively small, further studies with larger cohorts of healthy Mexicans are needed to determine if the frequencies of anti-DFS autoantibodies, detected by several confirmatory assays, are indeed higher than those previously reported in HI from other ethnic groups. If this turns out to be true, then it would be of interest to determine if there are any unique genetic, lifestyle, environmental, or inflammatory disease-related factors in Mexican and other Hispanic/Latino populations that might be associated with these elevated frequencies.