Page 28 - Read Online

Page 8 of 15 Zhou et al. J Cancer Metastasis Treat 2018;4:41 I http://dx.doi.org/10.20517/2394-4722.2018.16

The monosomy-7 cell remained slow-growing under both in vivo (as shown in Figure 3A by median 75 days
survival of mice with i.c. xenografts containing 45% monosomy-7 cells vs. 33 days survival of mice with i.c.
xenografts containing 20% monosomy-7 cells), and in vitro environments, and never became a population
larger than 5% in both U251 and U251-NS in vitro cultures, as well as in single-cell or low-density cultures
of U251. Thus, it would be the increase of Chr7-MS by STIC, not the increase of monosomy-7 cell growth
speed that explains the dramatic difference in increase of monosomy-7 cell percentage in xenografts from a
small number of cell implantation, as compared with that from 10 and 100-folds higher inoculum sizes. This
demonstrates the negative association of cell density and Chr7-MS rate by STIC in initial and subsequent cell
divisions following i.c. tumor cell implantation. The significantly higher percentage of TMC in xenografts
of U251NS-EFEMP1 (-Dox) with inoculum size of 100,000 cells compared with that of 10,000 cells is
functionally related to the shorter survival of mice from the fast growth features of TMC, although their
differences on Shannon diversity index and survival are not significant, but both are significantly different
from that of inoculum size of 1000 cells. Such cell density-related threshold of extracellular factors in control
of Chr7-MS rate were also observed in TMC in U251 in vitro culture under SA-conditions [Figure 2], both
demonstrating extracellular control of CIN in maintenance of TH.

EFEMP1 is an inhibitor of CIN
The cell density-dependent negative effect on CIN rate suggests paracrine-control of CIN. Below we present
the CIN inhibition function of an extracellular matrix protein EFEMP1 (also known as fibulin-3) that
was initially reported as a senescent protein , and later widely reported in cancers , with cell-context-
[37]
[36]
dependent dual functions in TMC and STIC in U251 and U251-NS lines, respectively .
[32]
Ectopic EFEMP1 was induced by adding Dox (1 mg/mL) to culture medium for about 1 day and maintaining
EFEMP1 overexpression in xenografts was achieved by providing Dox (1 mg/mL) in drinking water of
mice throughout the experiment. The tumor-promoting role of EFEMP1 in STIC, as suggested by its pro-
invasive function in STIC shown in a matrigel invasion assay, could only be seen in small inoculum sizes
of 1000 where the size of TMC number was too small to manifest EFEMP1’s suppression role, as shown
in xenografts from medium (10,000) and large (100,000) inoculum sizes . FISH analyses showed lack of
[32]
significant difference in both the steady state Chr7 subpopulations and H-index in xenografts of U251NS-
EFEMP1 (+Dox) of various inoculum sizes [Figure 3B], which was in striking contrast to that of U251NS-
EFEMP1 (-Dox) shown in Figure 3A. Besides the dual functions of EFEMP1 in TMC and STIC, EFEMP1
was further demonstrated to carry a role as InCIN, to suppress the increase of Chr7-MS by STIC during
formation of i.c. xenografts.

As reported previously in our studies of the tumor suppression function of EFEMP1 in glioma, long-time
in vitro overexpression of EFEMP1 in U251 amplified a population carrying two normal copies, denoted
as 2-Chr7 (2n), barely seen in parental culture, into the majority subpopulation (about 80%) in U251-
EFEMP1 (+Dox). In contrast to high tumorigenicity of U251 where TMC (3-Chr7 (2n, 1d)) was the dominant
subpopulation, U251-EFEMP1 (+Dox) with majority cells carrying 2-Chr7 (2n) showed significantly lower
tumorigenicity even after withdrawal of Dox in subcutaneous xenograft models . In this study, we examined
[13]
the effect of EFEMP1 on control of Chr7-MS rate in 2-Chr7 (2n) cells enriched in U251-EFEMP1 (+Dox).

FISH analysis was carried out on in vitro cultures of U251 transduced with the empty vector of pTRIPZ and
Dox-controlled transient- and stable-expression of ectopic EFEMP1. As shown in Figure 4, Chr7-defined
steady state of TH in U251 was similarly shown in U251-Vector after a 10-day Dox-treatment. The arrowhead
marked one chromosome 7 with 7q deletion (1d), which was specifically found in both TMC (2n, 1d) and
STIC (1n, 1d), as reported previously . For studying the effect of EFEMP1, U251-Vector was used as control
[13]
for the effect of vector and dox-treatment. As shown in Figure 4, Chr7-defined steady state of TH in U251
was dramatically altered due to EFEMP1 overexpression, with 69%, 44%, 9% of TMC present in cultures