| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Letters to the Editor |
Chain: A Novel Tumor-associated Transmembrane Protein in Primary Explants of Human Malignant Gliomas. Cancer Res., 60: 11681172, 2000.
The Milton S. Hershey Medical Center Hershey, Pennsylvania
I would like to address several issues with regard to an article by Joshi et al. (1)
published recently in Cancer Research: (a) the report cites the number of normal brain tissues evaluated for the capacity to bind radiolabeled IL1
-13 as assessed by autoradiography in error. Our group at a meeting (their Ref. 10
) reported these data. Shortly thereafter, these data documenting the absence of readily detectable binding of radiolabeled IL-13 collectively in more than a dozen normal brain tissue specimens were published (2
, 3)
; and (b) it should be noted that autoradiography is not sufficiently sensitive to detect IL-13 binding sites that are present at less than
400500 sites/cell. Our studies were thus aimed at showing a significant quantitative difference between normal brain and malignant glioma specimens with regard to the binding of IL-13 in situ. In particular, we were interested in binding sites that are not competed for by IL-4. Our more recent studies have shown, as expected, that the physiological receptor for IL-13, which is shared with IL-4 and is the only one known to transduce cellular signals, is expressed by central nervous system tissues and many vital peripheral organs (4)
. In unpublished findings, we have observed that the binding of IL-13 to some of these vital organs can be detected by autoradiography and it is competed for by IL-4.2
In distinct contrast to the broad expression of the IL-4-dependent IL-13 receptor complex, the receptor that is malignant glioma-associated (IL-13R
) has been found to be expressed in a very restrictive manner among normal tissues and to be present in malignant gliomas (4)
. Furthermore, specific binding of IL-4 to malignant glioma specimens was observed to be relatively weak or absent in situ (3)
.
These findings on expression of the shared physiological IL-13/IL-4 receptor on normal tissues and on malignant glioma tissues stood in contrast to a previous report that had failed to find the IL-13/IL-4 receptor on normal brain tissues (5) . In this latter report, IL-4 receptor could not be detected in five of six samples of normal human brain using a very sensitive reverse transcription-PCR assay with a very weak presence detected in the remaining sample. The unfortunate aspect of this report was that it might have provided a sense of undeserved safety in the justification for using an IL-4-based cytotoxin in a clinical trial in brain tumor patients (5) . Our data would provide no strong support for using IL-4 cytotoxin therapy for a tumor that does not overexpress the IL-4 receptor in situ and that is growing within normal tissue that clearly possesses it. Fortunately, in their most recent report, Joshi et al. (1) now demonstrate a prominent presence of the shared IL-13/IL-4 receptor expression within normal brain and in normal cells, at levels that are quite comparable with that seen in malignant tissues. This most recent result is in line with previous research in this area of investigation (6) and also with our findings (4) , although Joshi et al. (1) did not explain an apparent and important discrepancy in their findings.
Our research has consistently supported the concepts that (a) malignant gliomas are attractive tumors for targeted therapies using genetically engineered IL-13 as a delivery vector (2)
; and (b) there is a critical need to provide more selective delivery vectors based on dissimilarities between the glioma-associated receptor for IL-13 that is IL-4 independent and its physiological counterpart that is shared with IL-4. Several strategies have been outlined to satisfy the latter need (7
, 8)
. Substantial insight into the structure-function relationship of IL-13 and its receptors has been gained and we are exploiting this knowledge for the design of such new, unique delivery vectors (9
, 10)
. The most recent findings of Joshi et al. (1)
have served to emphasize and reconfirm both the attractiveness and the necessity to target an IL-4-independent receptor for IL-13 (IL13R
) for the treatment of malignant gliomas to obtain truly specific molecular targeting of this incurable malignancy. Unexpectedly, it seems now that such specific molecular recognition and targeting may be possible and applicable to a vast majority of patients with malignant gliomas (2
, 7)
.
Note Added in Proof
It seems incompatible that after having shown no major difference in the expression of IL4 receptor between normal and malignant brain (Cancer Res. 60: 11681172, 2000), Dr. Puri et al.published a subsequent report in which the claims of "Human glioblastoma but not normal brain cells express numerous receptors for the cytokine IL-4." and "Human brain tumor cells are particularly enriched in the expression of IL-4Rs on their cell surface" were reborn (Clin. Cancer Res. 6: 21572165, 2000).
FOOTNOTES
1 The abbreviation used is: IL, interleukin. ![]()
2 W. Debinski, B. Slagle-Webb, and D. Gibo, unpublished data. ![]()
The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
1 The abbreviations used are: IL, interleukin; IL-13R, interleukin 13 receptor; IL-4R, interleukin 4 receptor. ![]()
Received 3/24/01. Accepted 6/ 8/01.
REFERENCES
chain: a novel tumor-associated transmembrane protein in primary explants of human malignant gliomas. Cancer Res., 60: 1168-1172, 2000.Center for Biologics Evaluation and Research FDA, Bethesda, Maryland
Yale Medical School, New Haven, Connecticut
Center for Biologics Evaluation and Research FDA, Bethesda, Maryland
An issue was raised concerning our clear statement that normal brain tissues do not express detectable IL-13Rs1 (1) . This statement was based on our collaborative study with Dr. Debinski, and later he published a manuscript with same conclusions without our permission and knowledge (2) . On the basis of our new studies, however, we predict that IL-13 will bind to its receptors on normal brain tissues (1) . The technique used by Dr. Debinski is not appropriate for his conclusions (discussed below). Our laboratory first demonstrated the expression of high levels of high-affinity IL-13R on renal carcinoma cell lines (3) . Obviously, we were extremely interested in testing other cancer cell lines. As evident from various publications, we demonstrated that human glioma, AIDS-associated Kaposis sarcoma, prostate cancer, ovarian cancer cells, and normal skin fibroblasts also express varying numbers of IL-13Rs (4 5 6 7 8 9 10 11) . We also demonstrated that IL-13R could serve as a target for receptor-directed cancer therapy (4 5 7 , 9 , 10 , 12) . We continue to examine the complexities in IL-4 and IL-13 receptor interaction in various cell types, including cancer cells (13 , 14) . The study reporting differences in receptor structure between cancer cells and normal tissues as described in Joshi et al. (1) is one of the studies that unravels these differences.
The second issue raised concerns the expression of IL-13R in vital organs. It was already published in a cloning paper that IL-13R
' (also known as IL-13R
1) chain mRNA is widely expressed in almost all vital organs (15)
. We also reported that the IL-13R
' chain is expressed on normal brain tissue (1)
. However, we should be careful in interpreting Dr. Debinskis unpublished studies on binding in tissue sections, because normal tissues may harbor immune cells, such B cells and monocytes, and nonimmune cells, such as fibroblasts and endothelial cells, and these cells can bind IL-13 and IL-4. Therefore, binding and displacement characteristics of a ligand may not reflect only the properties of the normal tissue in question. Furthermore, the sensitivity of the assay also is questionable, as was also pointed by Dr. Debinski. Nevertheless, whether IL-13R forms a functional IL-13R complex in normal tissues is still not known and subject of ongoing investigations. Regarding IL-13R
(also known as IL-13
2; Ref. 16
), we have demonstrated for the first time that mRNA for the
2 chain is expressed on primary cell cultures derived from 82% of human glioma samples (1)
. In addition, we have reported that cell lines derived from ovarian carcinoma (8)
, skin fibroblasts (11)
, renal cell carcinoma (17)
, and pancreatic carcinoma express mRNA for the IL-13R
2 chain (18)
. A recent study has reported low-level expression of this chain in isolated glomeruli and cultured glomerular visceral epithelial cells (19)
. However, mRNA for this chain does not seem to be present on immune cells (14)
. In our study, we observed that mRNA for the IL-13R
2 chain is either expressed at low levels or not expressed at all in normal brain tissues (1)
. Our mRNA data were confirmed by immunofluorescence studies using commercially available antibody to the IL-13R
2 chain. Thus, the IL-13R
2 chain appears to be predominantly cancer-associated and is not only expressed in glioma cells but also on various cancer cell types. We are currently examining which other cancer cells express IL-13R
2 chain.
The third issue raised is the in situ expression of IL-4R on human glioma samples, which was not a main focus of the study by Joshi et al. (1)
. However, because the issue was raised, we will respond. We first reported that IL-4Rs are overexpressed on many solid cancer cells, including malignant gliomas (reviewed in Ref. 20
). Using surgical or biopsy samples, we have reported that the IL-4R
chain (also known as IL-4Rß) is expressed in
89% of glioblastoma multiforme samples as detected by reverse transcription-PCR (n = 36) and immunofluorescence (n = 11) studies (1
, 21)
. Because we and others have shown that IL-4R shares two chains with IL-13R, we examined the expression of various chains of IL-13R and IL-4R on brain tumor cells. IL-4Rs are comprised of two to three different types (21
, 22)
. In type I IL-4R, the IL-4R
chain forms a productive complex with the IL-2R
chain. In type II IL-4R, the IL-2R
chain is not present; instead, IL-4 binds to the IL-4R
chain and a novel protein that we later identified as an IL-13R
1 chain (22
, 23)
. We and others have also demonstrated that type II IL-4R is necessary for signal transduction induced by IL-4 or IL-13 (21
, 23
, 24)
. In our study of in situ expression of mRNA for IL-4R, we reported that five of six normal brain tissues did not express an IL-4R
(21)
. However, a study reported by Joshi et al. (1)
observed that total RNA derived from normal human astrocytes and a commercially available unidentified brain tissue expressed an IL-4R
chain. In fact, normal human astrocytes expressed a surface IL-4R
chain as detected by an immunofluorescence study. It is not an unexpected result, inasmuch as samples in our first study may have been obtained from areas of normal brain that did not express an IL-4R
chain. We have recently examined six additional samples (in addition to six reported in Ref. 21
) for the IL-4R
expression by reverse transcription-PCR, and we find that all six samples were weakly to moderately positive for IL-4R
chain mRNA. Thus 7 of 12 normal brain samples were positive for IL-4R
chain. Our observation was recently confirmed by Dr. Debinski himself in his own paper published in the May issue of Molecular Medicine (25)
. By Northern blot analysis, he finds that an IL-4R
chain was not expressed or was expressed at very low levels in cerebral cortex, occipital lobe, frontal lobe, temporal lobe, putamen, amygdala, caudate nucleus, corpus collosum, hypocampus, and whole brain. However, medulla, spinal cord, substantia nigra, and thalamus did express an IL-4R
chain, although at low levels detectable only after 2-week exposures of autoradiographic film. These faint bands could be long-term exposure artifacts, especially if the right cDNA probe was not used. Nevertheless, this study confirms our findings in Puri et al. (21)
that it is easy to miss certain areas of brain when biopsy samples are examined for IL-4R
mRNA expression. We would have very much liked if Dr. Debinski had performed Northern analysis or binding studies on cells obtained from primary samples of glioma tissues before suggesting that glioma cells do not express IL-4R. He did not show data that demonstrate that glioma cells do not express IL-4R. We have unequivocally demonstrated that IL-4Rs are expressed at the mRNA level in 32 of 36 primary samples of malignant glioma and at the protein level in 11 primary glioma samples. Finally, we would like to point out that radiolabeled binding studies on tissue sections are associated with many potential technical artifacts. In fact, we were not able to show specific binding of 125I-IL-13 on frozen sections of glioma tissues when similar techniques were used in collaboration study with Dr. Debinski (2)
.
The issue of IL-4 cytotoxin in clinical trials is beyond the scope of our paper (1) , and should have been raised in a different forum. In addition, Dr. Debinskis remark that our studies may have provided unsupported safety in the justification of use of IL-4 cytotoxin in the clinic for glioma therapy is unfounded and without any scientific merit and disregards all the preclinical studies that have been performed. As we have vigorously reported, IL-4 cytotoxin is not cytotoxic or is less cytotoxic to normal immune cells or endothelial cells as compared with cancer cells (20 , 26) . The differential sensitivity to normal and cancer cells is attributable to numerous reasons. Some of these reasons are: (a) low number of receptors; (b) affinity differences; (c) internalization rates; (d) intracellular processing; (e) metabolic rates; and (f) usage of various receptor chains. Dr. Debinskis assumption that some normal brain tissues expressing both chains of type II IL-4R will be sensitized to the cytotoxic effect of IL-4 toxin lacks scientific insight and vigor. Targeted cytotoxins require certain characteristics in a cell for cytotoxicity. Glioma cell lines seem to possess many of the above-mentioned characteristics and provide up to a 1000-fold therapeutic window in which one can target glioma cells but not the normal cells. On the basis of our numerous preclinical in vitro and in vivo studies (20 , 26) , we began a Phase I clinical trial, where IL-4 toxin has been infused directly into malignant brain tumors. The first study was published recently. We did not see appreciable evidence of IL-4 toxin-related toxicity to normal brain, although we infused up to 720 µg of IL-4 toxin. Extensive necrosis of malignant glioma was observed in six of nine patients (27) . Additional clinical studies are on-going using this targeted agent. Finally, based on numerous preclinical studies, we have also begun a Phase I clinical trial using IL-13-Pseudomonas exotoxin for the treatment of malignant glioma.
ACKNOWLEDGMENTS
Note Added in Proof
In contrast to what Dr. Debinski wrote in his Note Added in Proof, we have not directly compared the expression levels of IL-4R between normal brain and glioma cells. As stated in our response and also published (1), single normal human astrocytic cell line expresses mRNA and protein and several other normal brain tissues express low to moderate levels of mRNA for IL-4R
chain. Similarly, as stated in our previous papers, human glioblastoma cell lines are particularly enriched for the expression of IL-4R compared to normal immune and non-immune cells on which IL-4 has prominent biological activities.
Received 8/24/00. Accepted 5/10/01.
REFERENCES
chain: a novel tumor-associated antigen on malignant glioma cells. Cancer Res., 60: 1168-1172, 2000.
chain shared by receptors for interleukins 2, 4, 7, 9, and 15. J. Biol. Chem., 270: 8797-8804, 1995.
chain. J. Biol. Chem., 271: 29265-29270, 1996.
c chain of the IL-2 receptor. J. Biol. Chem., 272: 20251-20258, 1997.
chain and IL-4 induces the phosphorylation of JAK2 tyrosine kinase in human colon carcinoma cells. J. Biol. Chem., 270: 30829-30836, 1995.
' but not
chain: a functional component of interleukin-4 receptor. Blood, 91: 3884-3891, 1998.
c chain in RCC cells does not improve sensitivity. Cell. Immunol., 171: 80-86, 1996.[Medline]
This article has been cited by other articles:
![]() |
T. F. Liu, S. B. Tatter, M. C. Willingham, M. Yang, J. J. Hu, and A. E. Frankel Growth Factor Receptor Expression Varies among High-Grade Gliomas and Normal Brain: Epidermal Growth Factor Receptor Has Excellent Properties for Interstitial Fusion Protein Therapy Mol. Cancer Ther., August 1, 2003; 2(8): 783 - 787. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Cancer Research | Clinical Cancer Research |
| Cancer Epidemiology Biomarkers & Prevention | Molecular Cancer Therapeutics |
| Molecular Cancer Research | Cancer Prevention Research |
| Cancer Prevention Journals Portal | Cancer Reviews Online |
| Annual Meeting Education Book | Meeting Abstracts Online |