*2.5. LNC-RPS24, LNC-EPHA6 and LIPCAR Correlate with Soluble Thrombomodulin*

In order to analyze the association of lncRNAs with vascular injury due to AR, we studied the correlation of LNC-RPS24, LNC-EPHA6, and LIPCAR with vascular injury markers sTM and Ang-2 that were previously assessed [13]. There, we showed a significant increase of sTM levels in patients with acute rejection, followed by a subsequent normalization one year after transplantation, while the ratio between Ang-2 and Ang-1 (mainly determined by Ang-2) significantly increased during AR without significant changes afterwards. Here, no significant associations were found between LNC-RPS24, LNC-EPHA6, and LIPCAR with Ang-2. However, interestingly, LNC-RPS24, LNC-EPHA6, and LIPCAR correlated positively with sTM (Table 2).

**Table 2.** Correlation of lncRNAs with vascular injury markers sTM, Ang-2. Values represent correlation coefficient and *p*-value.


sTM = soluble thrombomodulin, Ang-2 = Angiopoietin-2.

## **3. Discussion**

Our study shows that levels of circulating LNC-EPHA6 are significantly higher in patients with T cell-mediated AR after renal transplantation, compared with kidney transplant recipients with a stable allograft function. LNC-EPHA6 remains elevated after AR, followed by a decrease one year after rejection. LIPCAR shows a similar pattern, but did not reach statistical significance. In addition, LNC-EPHA6, LIPCAR, and LNC-RPS24 correlate with the vascular injury marker sTM. This suggests that, in particular, LNC-EPHA6 may be related to microvascular damage, of which we confirmed its relation to AR by demonstrating a significantly lower presence of endothelial cells and pericytes in our renal biopsy study.

LNC-EPHA6 was earlier found to relate to diabetic nephropathy [19], but was not studied in the context of AR before. Our finding of higher LNC-EPHA6 levels in patients with AR compared with patients without AR provided proof of principle of the biomarker potential of lncRNAs in AR. However, here we analyzed four pre-selected lncRNAs, thus analyses of other lncRNAs in AR may yield additional associations and may potentially be important for prediction of (vascular injury after) AR. This is in line with two other studies that showed an association between lncRNAs and AR that suggested their value for diagnosis of AR in kidney transplantation [21,22]. Moreover, in a rat study, the lncRNA PRINS was shown to be significantly up-regulated in kidneys of rats with cold ischemia-elicited allograft rejection, compared with rats without rejection [23]. In addition, lncRNAs may also be of value in predicting the development of chronic damage after kidney transplantation [24].

Interestingly, levels of circulating LNC-EPHA6 decrease after AR, while eGFR remains stable. This substantiates that changes in LNC-EPHA6 are likely not to be related to changes in kidney function, but other factors in the pathogenesis of AR, such as persisting microvascular injury. This suggestion is supported by the strong correlation with sTM. However, although significant differences between immunosuppressive drug regimen were not observed, we cannot exclude that differences in rejection treatment altered levels of circulating lncRNAs. Furthermore, an association with the function of the EPHA6 gene might be possible, since lncRNAs are frequently co-regulated and co-expressed with their neighboring genes [25]. The EPHA6 gene is part of a EPH receptor tyrosine kinases family, and thereby interacts with ephrins which subsequently regulates several cellular processes including angiogenesis [26,27].

LIPCAR showed a similar trend after rejection as LNC-EPHA6. This could suggest a similar association as LNC-EPHA6 with rejection. However, changes in LIPCAR did not reach statistical significance due to a large variation. Analysis of LIPCAR in a larger cohort of patients with AR may confirm the link with vascular injury in rejection, since the size of our groups limits the interpretation of LIPCAR in our study. Circulating LNC-RPS24 was only marginally higher in rejection, but increased six months after year after rejection and remained higher. Although speculative, this may be the result of persistent vascular injury after AR or a consequence of the rejection treatment. Lastly, we found LncRNA MALAT1 to be only detectable in less than 30% of the patients in our cohort. Previously, MALAT1 was however detectable in most diabetes mellitus patients [19], suggesting that diabetes mellitus may increase circulating Malat1 levels. However, next to the previously mentioned limited group size, a relatively large spread of lncRNA levels within groups limits the possibility of drawing robust conclusions. The interpretation of subtle changes (e.g., correlation of lncRNAs with the specific Banff classification score for tubulitis, interstitial inflammation, and intimal arteritis) is difficult and larger groups are necessary for the identification of a specific lncRNA as a novel biomarker. However, differences in lncRNAs levels point out the interesting possible added value of lncRNAs in the context of acute cellular rejection. Identification of lncRNAs in the context of antibody-mediated rejection would be interesting as well, since this rare condition also has major implications for the amount of vascular injury.

In conclusion, LNC-EPHA6 is higher in kidney transplant recipients with rejection, compared with those without. This is the first study that shows changes in vascular injury related lncRNAs the first year after rejection. The results suggest that lncRNAs may reflect (micro)vascular damage in the context of rejection and emphasizes the potential role of lncRNAs as biomarkers to monitor vascular injury in kidney transplant rejection.
