Lung cancer is the leading cause of cancer death in men and women, both in the U.S. and worldwide, causing deaths – more than breast, prostate, and colorectal cancers combined. Five-year overall survival remains still less than 20% underscoring the need for a revolution in the management of patients with lung cancer. Although a series of key genes have been reported as playing a role in the tumorigenesis of lung cancer, over 40% of NSCLC patients carry no known mutation or clearly targeted therapeutic indication. And while surgical resection remains a mainstay of therapy, recurrence after surgery remains a serious problem even in stage I patients. We hypothesized that characterization of a new molecular marker with significant overexpression in NSCLC compared to matched normal tissues could identify a novel therapeutic or diagnostic marker. The work reported here describes a potential oncogenic role for Gremlin. To begin, we conducted a systematic review of published Oncomine data to identify tumor types in which Gremlin may be upregulated in comparison to normal tissue. We identified 24 datasets published to Oncomine that contained both tumor and normal (+)-JQ1 distributor samples suitable for analysis. Which it was down regulated. Of the datasets in which Gremlin was upregulated, four datasets were of lung tumors. Significant upregulation in four lung tumor datasets demonstrated the potential for a significant role of Gremlin in lung cancer. Following analysis of published microarray data to identify that Gremlin is significantly overexpressed in NSCLC microarray data, we performed quantitative RT-PCR analysis in our 96 AD with matched normal and 65 SCC with matched normal samples. Gremlin expression is significantly increased in lung AD samples compared to matched normal tissues. This is consistent with publicly available microarray data that Gremlin is highly overexpressed in tumor. Unlike two published SCC microarray datasets, we did not see a significant upregulation of SCC compared to matched normal tissue. It is possible that either sample size or a lack of paired samples in published microarray data may contribute to these findings. While our study utilized 65 pairs of SCC and matched normal tissues, samples from Garber et al and Bhattacharjee et al appear to have used 13 and 21 tumor specimens, respectively, but expression levels are compared with normal specimen expression levels pooled from all lung tumor subtypes included in the respective studies. The other possibility is a difference of platform. We conducted a quantitative RT-PCR analysis using a Taqman system while comparative studies utilized a microarray platform. Next, we checked the protein expression of Gremlin by IHC. We analyzed 24 pairs of AD and 8 pairs of SCC. As shown in Figure 3, Gremlin expression was observed in 67% of the AD samples to be greater than in matched normal slides. In SCC, only 2 of 8 samples demonstrated increased immunoreactivity in tumor specimens. We conducted a series of in vitro experiments with Gremlin transfection to assess the effect of Gremlin overexpression in lung fibroblast and epithelial cells. Transfection of GREM1 significantly increased cell growth in both normal lung fibroblast and epithelial cells. Moreover, many colonies were found in GREM1-transfected cells while almost no or low numbers of colonies were found in vectortransfected cells. The use of two additional independent cell lines – HLF-1 and NL-20 reiterated these findings. These in vitro data support our hypothesis that overexpression of Gremlin in lung cancer is involved in lung tumorigenesis and promotes cell growth and proliferation.