Endoplasmic reticulum heat shock protein gp96/grp94 is a pro-oncogenic chaperone not a tumor suppressor

To the Editor

We read with great interest the article by Chen et al. published in Hepatology (1). gp96, or grp94, is an endoplasmic reticulum (ER) HSP90 that chaperones many strategically important clienteles in oncogenesis including integrin, Toll-like receptor, Wnt co-receptor and insulin-like growth factor. However, Chen et al. demonstrated that conditional deletion of hsp90b1 (encoding gp96) and Pten from mouse livers using Cre recombinase driven by Albumin promoter (Alb-Cre) increased liver tumorigenesis (1). Nevertheless, the gp96 status of these tumors was not reported, leaving open a possibility of gp96 being either a tumor suppressor (if tumors were gp96−) or a pro-oncogenic chaperone (if gp96+). This diametrically opposing viewpoint is important to clarify particularly because liver-specific knockout of pro-tumorigenic genes could result in paradoxical enhancement of tumorigenesis from cells that escaped cre-mediated deletion (2). Herein, by crossing the same Alb-cre mice with our independently generated hsp90b1^flox/flox mice (3, 4), we also probed the roles of gp96 in liver oncogenesis. We found that knockout (KO) mice were more susceptible to hepatocyte carcinogenesis than wild type (WT) mice in response to diethyl-nitrosoamine (DENA), but the developing tumors were exclusively gp96+, due to an age-dependent expansion of the residual WT hepatocytes (Figure 1).

Figure 1. Alb-cre-gp96^flox/flox mice (KO) have increased liver carcinogenesis from residual gp96+ hepatocytes.

(A) Representative images of WT and KO tumors. (B) Quantification of liver tumors from WT and KO mice 62 weeks post DENA injection.

The development of gp96+ but not gp96− tumors in the same hosts illustrated its tumor-promoting rather than tumor-suppressive role. It might be prudent to examine if the Pten-loss tumors in the study of Chen et al. were also gp96+, analogous to another study on ER chaperone grp78 (5). It was found that liver cancer develops from escaping residual grp78+ hepatocytes after intended biallelic liver-specific deletion of Pten and grp78 via Albumin-Cre. The shared but apparently non-redundant cellular function of grp78 and gp96 in liver cancer underscores the critical oncogenic roles of both chaperones.