Cancer Diagnosis and Therapy
Expression of the human cachexia-associated protein (HCAP) in prostate cancer and in a prostate cancer animal model of cachexia
Article first published online: 25 FEB 2003
Copyright © 2003 Wiley-Liss, Inc.
International Journal of Cancer
Volume 105, Issue 1, pages 123–129, 20 May 2003
How to Cite
Wang, Z., Corey, E., Hass, G. M., Higano, C. S., True, L. D., Wallace, D., Tisdale, M. J. and Vessella, R. L. (2003), Expression of the human cachexia-associated protein (HCAP) in prostate cancer and in a prostate cancer animal model of cachexia. Int. J. Cancer, 105: 123–129. doi: 10.1002/ijc.11035
- Issue published online: 27 MAR 2003
- Article first published online: 25 FEB 2003
- Manuscript Accepted: 22 NOV 2002
- Manuscript Revised: 28 OCT 2002
- Manuscript Received: 22 JUL 2002
- VA/Department of Defense
- Richard M. Lucas foundation
- human cachexia-associated factor;
- prostate cancer (CaP);
Prostate cancer (CaP) patients with disseminated disease often suffer from severe cachexia, which contributes to mortality in advanced cancer. Human cachexia-associated protein (HCAP) was recently identified from a breast cancer library based on the available 20-amino acid sequence of proteolysis-inducing factor (PIF), which is a highly active cachectic factor isolated from mouse colon adenocarcinoma MAC16. Herein, we investigated the expression of HCAP in CaP and its potential involvement in CaP-associated cachexia. HCAP mRNA was detected in CaP cell lines, in primary CaP tissues and in its osseous metastases. In situ hybridization showed HCAP mRNA to be localized only in the epithelial cells in CaP tissues, in the metastatic foci in bone, liver and lymph node, but not in the stromal cells or in normal prostate tissues. HCAP protein was detected in 9 of 14 CaP metastases but not in normal prostate tissues from cadaveric donors or patients with organ-confined tumors. Our Western blot analysis revealed that HCAP was present in 9 of 19 urine specimens from cachectic CaP patients but not in 19 urine samples of noncachectic patients. HCAP mRNA and protein were also detected in LuCaP 35 and PC-3M xenografts from our cachectic animal models. Our results demonstrated that human CaP cells express HCAP and the expression of HCAP is associated with the progression of CaP and the development of CaP cachexia. © 2003 Wiley-Liss, Inc.
Cachexia is a severe debilitating condition characterized by anorexia, anemia, depletion of both fat and muscle tissues, hypoglycemia and asthenia involving cardiac and respiratory muscle as well as involving psychological distress.1, 2 Patients with cachexia often have a reduced response to chemotherapies and are more susceptible to secondary pathologies and therefore a shorter survival time.3, 4, 5 Approximately 70% of prostate cancer (CaP) patients with advanced disease suffer from cachexia, one of the most frequent causes of cancer-related death. Anorexia and increased energy expenditure cannot explain the entire syndrome since nutritional support is unable to reverse the process of cachexia in cancer patients.6, 7 The tumor may have both direct and indirect effects on the abnormalities associated with fat, carbohydrate and protein metabolism.8, 9 However, cachexia is not related to the anatomic site of the tumor, tumor cell type or tumor growth rate.10, 11 The molecular mechanisms and pathways responsible for cachexia are poorly understood because identification of the critical factors that cause cachexia remains elusive although there are many candidates.
Cytokines (TNFα, IL-6 and INFγ), neuropeptides (leptin and neuropeptide-Y), PTHrP and the transcription factor nuclear factor-kappa B (NF-κB) have been implicated in the development of cancer cachexia.12, 13, 14, 15, 16 In addition, increased attention has been drawn to putative tumor-derived cachectic factors. Murine lipid-mobilizing factor (LMF) and its human homologue zinc α-2-glycoprotien (ZAG) were associated with catabolism of adipose tissue in cachexia.17, 18 Anemia-inducing factor (AIF) was found to be responsible for the development of anemia and immunodeficiency observed in cancer cachexia.19, 20 Muscle wasting is considered a major contributor to the morbidity and mortality in cancer. Tisdale et al.1 observed that a highly active murine proteolysis-inducing factor (PIF) was directly involved in muscle protein degradation using both in vivo and in vitro models. PIF was detected in the urine of patients with cancer-associated cachexia, including those of colon, lung and pancreas.21, 22, 23, 24, 25 Recent studies by Watchorn et al.26 showed that PIF is also expressed in tissues other than skeletal muscle, such as liver. PIF differentially induces activation of NFκB and signal transducer and activator of transcription (STAT) 3, resulting in the induction of proinflammatory cytokines (IL-8 and IL-6) and induces the shedding of syndecans from the cell surface. The authors propose that PIF, in addition to its role in cachexia, may be involved in embryologic development. A possible human homologue of PIF, human cachexia-associated protein (HCAP), was identified in a breast cancer library by a homology search of the available 20-amino acid sequence of PIF.27 The sequence of HCAP has a 90% homology to PIF and is absent from libraries of matched nontumor samples (Genebank accession number AR053250).
In our study, we have investigated the expression of HCAP message and protein in human CaP cell lines, CaP tissues, CaP metastases and in CaP xenografts that induce body weight loss in mice. In addition, we tested for the presence of HCAP in the urine from cachectic and noncachectic CaP patients.
MATERIAL AND METHODS
The following human CaP cell lines were used in the studies: PC3, LNCaP, DU145 (obtained from ATCC, Rockville, MD), B-PC3 (invasive), N-PC3 (noninvasive) (sublines of PC-3, gift from Dr. Stearns, Medical College of Pennsylvania, PA), PC-3M (metastatic subline of PC-3, gift from Dr. Pettaway, University of Texas M.D. Anderson Cancer Center, Houston, TX), C4-2 and C4-2B (sublines of LNCaP from Urocor, Oklahoma City, OK). A fetal foreskin fibroblast primary culture established in our laboratory was used as a control. All of the cell lines with the exception of PC-3M were cultured in RPMI-1640 medium (Life Technologies, Gaithersburg, MD) supplemented with 10% fetal bovine serum (FBS). The PC-3M cells were cultured in RPMI-1640 medium with 2 mM glutamine and 10% FBS.
All donors and/or families signed informed consent that was approved by the University of Washington division of Human Subjects.
Nonmalignant prostate specimens, primary tumors and metastatic tissues were obtained either from radical prostatectomies or from rapid autopsies, i.e., tissue acquisition within 3 hr of death. Normal prostate tissue was obtained at the time of cadaveric-organ procurement for transplantation. Tissue samples were either frozen in liquid nitrogen or fixed in 10% buffered neutral formalin and subsequently embedded in paraffin. Bone metastases were demineralized in formic acid prior to paraffin embedding.
Urine from 28 advanced CaP patients and 10 normal subjects were collected and frozen at −20°C until studied. Normal subjects were weight stable, and the cancer patients were either weight stable or showed various degrees of weight loss. A patient was considered to be cachectic when loss of both weight and appetite (anorexia) was recorded in the patient chart. No patient was receiving chemotherapy or radiation therapy at the time of study.
RNA isolation and reverse transcriptase polymerase chain reaction (RT-PCR)
Total RNA was extracted from tissues and cell lines using RNA-STAT 60 (Tel-test, Friendswood, TX) according to the manufacturer's protocol. One microgram of total RNA was used for reverse transcription using the 1st-Strand™ cDNA synthesis kit (Clontech, Palo Alto, CA). β2-microglobulin (MIC) PCR was used as a control for RNA quality and RT performance. Sense and antisense primers for HCAP were ACTCTCCTCTTCCTGACAGCTCTGG and CTGCTGCTCCTGGGTATCATTTCTC and for MIC were CACGTCATCCAGCAGAGAATGGAAAGTC and TGACCAAGATGTTGATGTTGGATAAGAG. PCRs for both messages were performed using a 2-step procedure with the following conditions: 3 min at 80°C for 1 cycle, 5 sec at 95°C, 1 min at 69°C for 35 cycles and 7 min at 72°C for 1 cycle. PCR products for HCAP were purified by gel electrophoresis (QIAquick Gel Extraction Kit, Qiagen, Valencia, CA) and sequenced (Center for AIDS Research DNA Sequencing facility at the University of Washington).
Generation of RNA probes
The 361 bp HCAP PCR amplicon was cloned into the pGEM-T vector (Promega, Madison, WI). The digoxigenin-labeled sense (negative control) and antisense probes were synthesized using the DIG RNA labeling kit (Roche Molecular Biochemicals, Indianapolis, IN) with T7 or SP6 RNA polymerase, respectively.
In situ hybridization (ISH)
Six-micrometer-thick sections of formalin-fixed, paraffin-embedded tissue were mounted on Superfrost Plus slides (VWR Scientific, West Chester, PA). Slides were baked at 60°C for 1 hr, deparaffinized in 3 changes of xylene followed by dehydration in 80%, 95% and 100% ethanol. Hybridizations were performed using the Ventana genII automated ISH system (Ventana Medical System, Tucson, AZ) according to the manufacturer's protocol. Tissue RNA denaturation was done at 80°C. Fifteen nanograms of the RNA probes were added manually and hybridized at 48°C for 5 hr. Sections were washed at 40°C with 1 × 2X and 2 × 0.5X SSC (sodium chloride/sodium citrate). Bound probes were detected with antidigoxigenin antibody and visualized using DAB as a substrate.
Protein preparation from tissues and urine
Tissue samples (100 mg) were used for protein extraction. Five hundred microliters of lysis buffer (50 mM Tris-HCL, pH 7.5, 150 mM NaCl, 1% Nonidet P-40, 0.5% sodium deoxycholate and complete protease inhibitors, Roche, Indianapolis, IN) were added to the sample, and homogenization was carried out using the OMNI TH Homogenizer (OMNI International, Warrenton, VA). Samples were centrifuged at 16,000g for 10 min and the supernatant was collected. The protein concentration was determined using the Bio-Rad DC protein assay (Bio-Rad, Hercules, CA).
Aliquots of the collected urine samples (100 ml) were centrifuged at 5,000g at 4°C for 30 min to remove debris. Proteins were precipitated with 80% (w/v) ammonium sulfate overnight at 4°C and recovered by centrifugation at 19,600g at 4°C for 30 min. The protein precipitate was dissolved in PBS and dialyzed against PBS overnight at 4°C, followed by concentration using an Ultrafree-15 Centrifugal Filter with a molecular weight cut-off of 10 kDa (Millipore, Bedford, MA). The protein concentration was determined as above.
Generation of anti-HCAP antibody
A peptide consisting of the 25 amino acids (YDPEAASAPGSGNPCHEASAAQKEN) at the N-terminus of the deduced HCAP protein that overlapped with the known 20-amino acid sequence of PIF was coupled to keyhole limpet hemocyanin (KLH) and used to generate rabbit antisera (SynPep, Dublin, CA). Pre-immune sera were collected as controls.
Fifteen micrograms of total protein were separated under reducing conditions on 15% SDS Ready Gels (Bio-Rad), and transferred onto nitrocellulose membranes (Bio-Rad). Membranes were blocked with 5% nonfat milk, 0.1% Tween-PBS (w/v) overnight at 4°C and then incubated with either 10 μg/ml mouse monoclonal anti-PIF antibody or 1:5,000 diluted rabbit anti-HCAP serum in 0.1% Tween-PBS for 1 hr at room temperature. Rabbit anti-mouse IgG1 or rabbit pre-immune serum was used as a negative control in the experiments involving rabbit anti-HCAP. Horseradish peroxidase (HRP)-labeled goat anti-mouse (1:4,000) (Amersham, Piscataway, NJ) and goat anti-rabbit (1:1,000)(Santa Cruz, Santa Cruz, CA) secondary antibodies were used to detect bound anti-PIF and anti-HCAP antibodies, respectively, for 1 hr at room temperature. Immunoreactive proteins were visualized using the ECL system (Amersham).
Immunoprecipitation of HCAP protein from tissues and urine
Total IgG was affinity purified from the rabbit anti-HCAP antisera using Protein A Sepharose (Zymed, South San Francisco, CA). The IgG was bound and cross-linked to immobilized protein A using the Seize X Protein A Immunoprecipitation Kit (Pierce, Rockford, IL). Five-hundred micrograms of total protein from the urine samples or metastatic tissues were applied onto the antibody-protein A resin, and the protein was immunoprecipitated according to the manufacturer's protocol.
PC-3M and LuCaP35 cachectic animal models
All animal procedures were approved by the University of Washington Animal Care Committee.
Three groups of animals were used: (i) control animals without tumor; (ii) animals bearing PC-3M xenografts; and (iii) animals bearing LuCaP 35 xenografts. Eight-week-old male SCID mice (Charles-River Laboratory, Wilmington, MA) were implanted subcutaneously in the shoulder with either 2.5 × 106 PC-3M cells, a PC-3 metastatic variant or 25 mm3 fragments of LuCaP 35 tumor.28 PBS was injected in the negative control group. Eleven mice were used for each group. The weights of all animals were recorded at the start of the experiment and were monitored every other day thereafter. Food intake was checked weekly. Tumors were measured 3 times a week and converted to tumor volume using the formula length × width × height × 0.5236. Mice were anesthetized with ketamine/xylazine (130 mg/8.8 mg/kg body weight) and sacrificed by cervical dislocation. No mice were allowed to exceed a 20% weight loss or a tumor size >1,000 mg. Tumor samples were dissected and processed as described in the Tissues section. Blood was collected using cardiac puncture, and the serum was removed after centrifugation at 6,600g for 10 min. The levels of serum triglycerides were measured using Infinity Triglyceride Reagent (Sigma, St. Louis, MO).
All statistical and correlation analyses were performed using PRISM TM version 2.0 (GraphPad, San Diego, CA). Significance of differences between groups was determined using Fisher's exact test, and correlations were calculated using the Spearman test.
Expression of HCAP mRNA by RT-PCR
The 361 bp HCAP PCR amplicon was detected in all of the CaP cell lines (Fig. 1) except PC-3M. The HCAP mRNA was not detected in fetal foreskin fibroblasts. The PCR amplicons from each cell line were sequenced and shown to be 100% identical to that of HCAP.
The expression of HCAP mRNA in prostatic tissues was determined by RT-PCR using samples of 5 normal prostates (cadaveric-donor transplant program), 2 primary tumors obtained at rapid autopsies, 15 matched pairs of nonmalignant and neoplastic tissues from radical prostatectomies from patients with organ-confined disease (for representative data, see Figure 2a) and 9 bone metastases (Fig. 2b). The presence of HCAP mRNA was observed in both of the primary CaP tissues and in 13 of 15 radical prostatectomy cancer specimens and 7 of 9 bone metastases but not in normal prostate or any adjacent, nonmalignant regions of the prostates removed at the time of surgery or rapid autopsy (Fig. 2c).
Localization of HCAP mRNA expression by ISH
ISH was used to determine the cellular source of HCAP mRNA in early and advanced disease CaP specimens. A high level of staining was observed in the cancer cells in 2 of 2 primary tumors removed at radical prostatectomy and in 3 of 3 primary tumors obtained from rapid autopsies (Fig. 3). Note that the adjacent nonmalignant prostate epithelial cells and the stromal cells were not stained. Four osseous metastases, 2 lymph node metastases and 2 liver metastases also exhibited expression of HCAP in the metastatic foci with no staining in the adjacent bone matrix or stroma (Fig. 3). The control slides with sense probe were negative.
Detection of HCAP protein in CaP by Western blot
The rabbit anti-HCAP antisera was tested and shown to detect recombinant HCAP (gift from Drs. Jeremy Graff and James Miller, Eli Lilly, Indianapolis, IN) on a Western blot (data not shown). Since we detected HCAP mRNA in CaP specimens, we next examined whether CaP cells express the HCAP protein. HCAP protein was detected in 4 of 5 bone, 3 of 5 liver and 2 of 4 lymph node metastases (representative data are shown in Figure 4a). The protein was absent in the normal prostate tissue from 5 cadaveric organ donors (Fig. 4b). It was also absent in 4 pairs (malignant and nonmalignant sections) of prostatic tissues from radical prostatectomies on patients with organ-confined disease (data not shown).
Detection of HCAP protein in the urine of CaP patients by Western blot
Urine samples from cachectic (n = 19) and weight-stable CaP patients (n = 9) along with normal subjects (n = 10) were collected to test for the presence of HCAP protein and to determine whether its presence is associated with CaP cachexia. Representative data are shown in Figure 5. The 24 kDa HCAP protein was detected in 9 of 19 of the samples from cachectic CaP patients using the rabbit anti-HCAP anti-sera, and an equivalent molecular weight band was detected using the monoclonal anti-PIF antibody (Fig. 5a,b). The protein was not detected in any samples from normal subjects and weight-stable CaP patients. These groups were statistically different from the cachectic CaP group (p = 0.0011) (Fig. 5a).
HCAP expression in PC-3M and LuCaP 35 xenografts in mice experiencing body weight loss
All 11 PC-3M-bearing mice experienced weight loss. They displayed a 15.1 ± 1.5% (mean ± SEM) weight loss after 26 days with a tumor volume of 180 ± 30 mm3. Mice bearing LuCaP 35 tumors experienced a 7.91 ± 1.8% average weight loss (n = 9; 2 of the 11 mice in this series did not complete the study) after 35 days with tumor volume of 592 ± 197 mm3 (Fig. 6). Three of the 9 did not lose weight, although the tumors were not different in size from those in the 6 mice with weight loss (11.8 ± 2.1%). A drop in the serum triglyceride level is one criterion indicating the presence of cachexia in animal models.23, 29 Both tumor groups showed a significant decrease (p < 0.0001) in serum triglyceride levels at the time of sacrifice compared with normal controls (Fig. 7).
We tested for the presence of HCAP mRNA and protein in the CaP xenografts. RT-PCR demonstrated the presence of HCAP mRNA in 8 of 11 PC-3M and 5 of 6 LuCaP 35 tumors from mice with weight loss. No HCAP mRNA was detected in 3 LuCaP 35 tumors from animals that did not lose weight. There was a significant correlation between animals that lost weight and the expression of HCAP mRNA in the tumors (r = 0.74 and 0.79 for PC-3M and LuCaP 35, respectively). HCAP protein was detected in PC-3M tumors by both anti-HCAP and anti-PIF antibodies on Western blot after protein immunoprecipitation with anti-HCAP antibody (Fig. 8). The HCAP protein was undetectable in LuCaP 35 tumors.
Cancer-associated cachexia accounts for a significant portion of the morbidity seen in advanced cancer patients and is probably directly or indirectly responsible for a substantial percentage of patient deaths. Despite its prevalence, the mediators of cachexia are ill-defined, although the literature suggests multiple candidates.2 A relatively new addition to this list is the tumor-derived factor PIF.22 Although the literature on PIF is not extensive and mostly consists of reports using murine models, there are several findings that would strongly suggest a role for PIF in cancer-associated cachexia. For example, PIF was initially discovered in the sera of mice bearing the cachexia-inducing MAC 16 tumor but not in the sera of the noncachectic MAC 13 tumors.30 Furthermore, cachexia could be induced in mice upon injection of the PIF-containing serum and the effect neutralized with anti-PIF antibodies. Although HCAP is believed to be the human homologue of PIF, the degree of homology and common identity of PIF and HCAP remains elusive since PIF has not been fully sequenced. A factor initially identified as PIF, due to its reactivity with anti-PIF antibodies, was found in the urine of patients with tumors of the colon, lung or pancreas,23, 24 and this factor was shown to result in muscle protein degradation.31, 32 In the work presented here, we have utilized the term HCAP to refer to the human cachexia-inducing factor that may or may not be identical to PIF. To our knowledge, there have been no reports in the literature that have attempted to explore the expression of HCAP in human CaP.
Using RT-PCR primers designed from the published sequence of HCAP, we observed that HCAP mRNA is expressed by CaP cell lines, primary tumors and metastases but not in normal prostate. ISH studies confirmed that the HCAP message was cancer cell associated and not present in the stroma, adjacent nonmalignant glands or normal bone marrow cells surrounding bone metastases. The association of HCAP with CaP was also shown at the protein level. For example, HCAP protein was detected by Western blot in tissue extracts of CaP metastases but not from homogenates of organ-confined disease. The fact that we failed to detect the protein in organ-confined tumors yet found HCAP message is not an uncommon occurrence and fits the clinical profile since cachexia is not associated with early disease. Some of the strongest HCAP protein bands on Westerns were observed in the urine of CaP patients who were cachectic (vida infra). As might be expected, identical appearing bands were detected with an anti-PIF antibody.
One of the primary clinical manifestations of cancer-associated cachexia is weight loss. The degree of weight loss varies among patients but correlates well with decreased survival.2 We questioned whether HCAP protein would be detected in the urine of patients with advanced disease with or without weight loss associated with a diagnosis of cachexia. Unfortunately, the presence of edema in many patients with advanced disease obscures the actual wasting of muscle and fat that results in true weight loss. Therefore, it was not possible in many of our patients to quantify the amount of weight loss or to directly correlate weight loss with the presence of HCAP protein in the urine. Nevertheless, there was a good correlation between the detection of HCAP protein in the urine and a diagnosis of cachexia in our study population compared with advanced-disease patients without a diagnosis of cachexia. We did not detect HCAP in urine in some patients with weight loss, and this is in agreement with observations reported by Cariuk et al.23 that PIF was undetectable in the urine of some patients experiencing weight loss, especially if the weight loss was <1 kg/month. A limitation of our studies and those in the past is that detection of HCAP or PIF by Western blot is not a highly sensitive technique, and the absence of a target band may be due to several factors including presence of the protein below the level of detection and/or degradation of the protein during the handling and processing of the urine. With the recent availability of recombinant HCAP, there will likely be development of immunoassays that should greatly increase the sensitivity for detecting the HCAP protein. Of course, it is also possible that weight loss in these particular patients may be caused by factors other than HCAP. For example, we recently observed that cytokines TNFα, IL-6 and IL-8 were elevated in cachectic prostate cancer patients compared with noncachectic, advanced-stage prostate cancer patients.33
Additional evidence that HCAP may be associated with CaP came from our efforts to develop suitable xenograft models for the study of cachexia. Both PC-3M and LuCaP 35 xenografts resulted in weight loss (approximately 15% and 8% at 1 month, respectively) when the tumor weight was subtracted from the body weight. In both xenografts, there was a good correlation between weight loss and the expression of HCAP message. Interestingly, PC-3M as a cell line does not express HCAP message. Studies are underway to determine how soon after implantation the HCAP message appears and how soon after excision and placement back into culture the message disappears. We were unable to determine whether the same scenario existed for LuCaP 35, since this xenograft is passaged serially in mice and does not grow as a cell line. Similar to our findings with the clinical specimens, the HCAP protein was also detected by Western blot in the PC-3M xenografts. However, we were not able to detect the HCAP protein in the LuCaP 35 xenografts. We attribute this finding to a low abundance, but not necessarily the absence, of HCAP protein in these xenografts. Recall that the weight loss in the LuCaP 35 series was less than that of the PC-3M xenografts despite being nearly 3 times the size (592 mm3vs. 180 mm3, respectively).
A limitation in our xenograft studies was the inability to test for proteolysis, such as a decrease in lean muscle mass. A drop in serum triglycerides as noted herein, although reportedly consistent with the onset of cachexia in animal models,23 is not a measure of proteolysis. To establish that HCAP is involved in cachexia, it will be necessary to associate it with a decrease in lean muscle mass or another indicator of proteolysis.
The relationship between PIF and HCAP both molecularly and in their functional role in cachexia has recently become more complicated as other reports in the literature reveal that the sequence of HCAP, sometimes apparently unbeknown to the authors, is highly homologous or identical to several newly identified molecules: 98% identical to the 40 amino acids of Y-P30, a survival-promoting peptide for neurons;34 including 98% identical to a survival-evasion peptide that can attenuate retinoic acid responses and protect cells35 (AY044239 and AC079310); 100% identical to the 458 bp Dermcidin (AF144011), an antimicrobial peptide secreted by sweat glands;36 and 100% identical to a preproteolysin gene (AF418981). The biologic significance of the homology between HCAP and these other molecules remains to be determined. The apparently high diversity in biologic function is certainly intriguing.
In conclusion, to our knowledge, this is the first study demonstrating an association between the cachectic factor HCAP and CaP. The association was observed in cell lines, metastatic tumors and in the urine of CaP patients with cachexia but not in the urine of those with advanced disease that were not cachectic. Protein of equivalent size from these positive urines was detected by Western blot using both anti-HCAP and anti-PIF antibodies. Furthermore, HCAP message but not protein was found in primary tumors. Finally, PC-3M and LuCaP 35 xenografts induced weight loss in mice consistent with cachexia and two-thirds of these tumors expressed HCAP. In the future, these models will become increasingly important as attempts are made to further define the mechanisms and pathways associated with cachexia, to elucidate the function of HCAP, to explore the targeting of HCAP and in due course, to test therapeutic strategies.
We thank Dr. A. Liu for his critical reading of the manuscript; Dr. M. Roudier for her valuable discussion, comments and clinical support; Dr. J. Pfitzenmaier for his assistance in statistical analysis; and Ms. J. Quinn, Ms. L. Brown, Mr. A. Odman, Mr. E. Afman and Mr. D. Felise for their technical support. We also thank Drs. J. Graff, B. Konicek, K. Kikly and J. Miller of Eli Lilly for consultation and ongoing collaboration. This work was supported by a VA/Department of Defense grant to RLV and the Richard M. Lucas foundation.
- 26Proteolysis-inducing factor differentially influences transcriptional regulation in endothelial subtypes. Am J Physiol Endocrinol Metab 2002; 282: 763–9., , .
- 27Human cachexia associated protein. Palo Alto, CA: Incyte Pharmaceuticals, Inc., 1998. 628413(5,834,192)., .
- 33Elevation of cytokine levels in cachectic prostate cancer patients. Cancer, in press., , , , , .