In the present study, we modified xylose uptake properties of a recombinant xylose-utilizing yeast Saccharomyces cerevisiae by expression of heterologous and homologous permease-encoding genes. In a mutant yeast strain with the main seven hexose transporter genes deleted, and engineered for xylose utilization, we screened an expression cDNA library of the filamentous fungus Trichoderma reesei (Hypocrea jecorina) for enhanced growth on xylose plates. One cDNA clone with significant homology to fungal sugar transporters was obtained, but when the clone was retransformed into the host, it did not support significant growth on xylose. However, during a long liquid culture of the strain carrying the cDNA clone, adaptive mutations apparently occurred in the host, which led to growth on xylose but not on glucose. The new transporter homologue, Trxlt1 thus appears to code for a protein specific for xylose uptake. In addition, xylose-transporting properties of some homologous hexose transporters were studied. All of them, i.e., Hxt1, Hxt2, Hxt4, and Hxt7 were capable of xylose uptake. Their affinities for xylose varied, K (m) values between 130 and 900 mM were observed. The single-Hxt strains showed a biphasic growth mode on xylose, alike the Trxlt1 harboring strain. The initial, slow growth was followed by a long lag and finally by exponential growth.

The serine-threonine protein phosphatase PPM1D is likely to play an important role in tumorigenesis. Through inactivation of p38 MAPK, PPM1D acts as a negative feedback regulator of p53 tumour suppressor gene and controls the expression of other cell cycle regulatory proteins, such as CCND1. In addition, recent knock-out mouse studies implicated PPM1D in the regulation of p16 expression and the RB tumour suppressor pathway. Here we explored the role of PPM1D aberrations in primary breast cancer. PPM1D copy number analysis showed amplification in 11% (13/117) of the tumours and quantitative real-time RT-PCR revealed a significant correlation (p = 0.0148) between PPM1D amplification and increased expression. PPM1D amplification occurred almost exclusively in tumours with wild-type p53 suggesting that these events are mutually exclusive and further confirming the role of PPM1D as a negative regulator of p53. Interestingly, PPM1D amplification was associated with ERBB2 expression (p = 0.0001) thus implying that PPM1D aberrations occurs in tumours with poor prognosis. We also explored the expression levels of two possible downstream targets of PPM1D. However, immunohistochemical analyses revealed no differences in the staining patterns of CCND1 and p16 proteins in tumours with or without PPM1D aberrations, thus suggesting that previous data from animal model experiments is not directly transferable to primary human tumours. On the other hand, these key cellular proteins are likely to be regulated through a complex fashion in breast cancer and apparently PPM1D represents only one of these mechanisms. Taken together, our findings substantiate an important role for PPM1D in breast cancer.

Bone morphogenetic proteins (BMP) make up a family of extracellular signaling molecules that play a critical role in vertebrate development and both inhibit and stimulate growth in cancer cells. BMP7 was recently identified in our genomewide copy number and expression survey as being activated through amplification in breast cancer cell lines. In the present study, we further explored BMP7 gene copy number and expression changes in 22 breast cancer cell lines and 146 primary breast tumors. FISH analysis revealed that BMP7 copy number varied greatly from one cell line to another, with three cell lines showing extremely high-level amplification. Among primary tumors, BMP7 copy number was increased in 16% of the cases. BMP7 mRNA expression was determined in the cell lines and in a subset of 44 tumor samples by RT-PCR or quantitative real-time RT-PCR, respectively. Despite elevated mRNA levels in cancer cells, there was no significant association between copy number increase and mRNA expression, even though the highest expression was seen in cell lines and tumors with increased BMP7 copy number. Most interestingly, immunohistochemical analysis revealed BMP7 protein staining in all 11 breast cancer cell lines examined and strongly elevated BMP7 protein expression in 71.4% of the tumor samples as compared to normal mammary epithelium. Our results illustrate the frequent involvement of BMP7 alterations in breast cancer and especially highlight overexpression of the BMP7 protein in a very large fraction of primary breast tumors, thus suggesting a possible functional role for BMP7 in breast cancer development.

In cell extracts all of the nonliganded steroid receptor molecules are found as an oligomeric complex with Hsp90 and other proteins. In previous studies we have shown that Wild-type Hsp90 and progesterone receptor (PR) are located in different cell compartments (Tuohimaa et al. [1993] Proc. Natl. Acad. Sci. USA 90:5848-5852). In the present work we studied whether PR and Hsp90 can efficiently associate provided they are present in the same cell compartment. The association of Hsp90 with PR in vivo was studied by nuclear cotranslocation and immunohistochemistry with an antibody (alphaD) which can distinguish between the oligomeric and dissociated form. Upon expression of a cytoplasmic mutant of PR with Wild-type (cytoplasmic) Hsp90 and Wild-type (nuclear) PR with NLS-Hsp90 (a Hsp90 with a nuclear localization signal), we noted that the epitope of alphaD in PR was exposed in both cases. Also, in vivo crosslinking and treatment of cells with substances which stabilize the oligomeric complex in vitro were inefficient in demonstrating or inducing a similar oligomeric receptor form detectable in vitro in cell homogenates. However, when the cytoplasmic PR mutant (DeltaPR) was coexpressed with a nuclear form of Hsp90 (NLS-Hsp90), a portion of PR was cotranslocated into the nucleus. This would indicate that steroid receptors are indeed associated with Hsp90 in intact cells, but the Hsp90-associated receptor pool represents only a small portion of the receptors. This suggests that the majority of oligomeric complexes seen in cell extracts are formed during cell fractionation.

In cell extracts all of the nonliganded steroid receptor molecules are found as an oligomeric complex with Hsp90 and other proteins. In previous studies we have shown that Wild-type Hsp90 and progesterone receptor (PR) are located in different cell compartments (Tuohimaa et al. [1993] Proc. Natl. Acad. Sci. USA 90:5848-5852). In the present work we studied whether PR and Hsp90 can efficiently associate provided they are present in the same cell compartment. The association of Hsp90 with PR in vivo was studied by nuclear cotranslocation and immunohistochemistry with an antibody (alphaD) which can distinguish between the oligomeric and dissociated form. Upon expression of a cytoplasmic mutant of PR with Wild-type (cytoplasmic) Hsp90 and Wild-type (nuclear) PR with NLS-Hsp90 (a Hsp90 with a nuclear localization signal), we noted that the epitope of alphaD in PR was exposed in both cases. Also, in vivo crosslinking and treatment of cells with substances which stabilize the oligomeric complex in vitro were inefficient in demonstrating or inducing a similar oligomeric receptor form detectable in vitro in cell homogenates. However, when the cytoplasmic PR mutant (DeltaPR) was coexpressed with a nuclear form of Hsp90 (NLS-Hsp90), a portion of PR was cotranslocated into the nucleus. This would indicate that steroid receptors are indeed associated with Hsp90 in intact cells, but the Hsp90-associated receptor pool represents only a small portion of the receptors. This suggests that the majority of oligomeric complexes seen in cell extracts are formed during cell fractionation.