PROJECT SUMMARY
It is increasingly recognized that certain metabolic enzymes are required in cells not for what they produce,
but instead for processing and thus preventing the accumulation of their substrates which may have toxic
properties. Such enzymes can be attractive therapeutic targets, as their inhibition can poison cancer cells
with self-produced toxic metabolites in a manner that is highly dependent on production of the toxic
metabolite. Here we investigate a new detoxifying enzyme, ketodehydrosphinganine reductase (KDSR),
which is part of the de novo sphingolipid biosynthesis pathway. We find that KDSR is not required to
provide sphingolipids, as cancer cells can readily salvage them from their environment, but instead is
needed to prevent accumulation of its substrate 3-ketodehydrosphinganine (3KDS). Accumulation of
3KDS, either via KDSR KO or by direct treatment of 3KDS to cells, appears to disrupt the endoplasmic
reticulum (ER) and cause an overload of misfolded proteins in cancer cells. This indicates KDSR as a
potential cancer therapy target capable of impairing ER function and proteostasis in cancer cells, which
we will explore in this proposal. In Aim 1, we will examine the upstream steps that drive 3KDS production,
which we hypothesize are elevated in multiple cancer subtypes, and thus directly renders the cells
dependent on KDSR for 3KDS detoxification. These will be further considered as possible biomarkers for
tumors that would respond to KDSR targeting. In Aim 2, we will examine how 3KDS accumulation disrupts
the ER and leads to death, and the responses mounted by cancer cells to counter 3KDS toxicity. In Aim
3, we will gauge the therapeutic potential of targeting KDSR by comparing 3KDS production capacity
between tumor tissues and normal tissues from animal models and from deidentified patient tissues. In
this manner we hope to provide a working blueprint for how to selectively target subtypes and
subpopulation of cancer cells based on their 3KDS producing activities, provide biomarkers which predict
whether a tumor will respond to such a therapy, and provide new insights into the endoplasmic reticulum-
and proteostasis- related vulnerabilities of cancer cells.