TO THE EDITOR:
We read with great interest the recent article by Hornick et al. (1) on the subcellular distribution of 111In- and 125I-labeled somatostatin analogs. We applaud the authors for addressing the important question of where radiolabeled somatostatin analogs localize inside tumor cells, because there is considerable clinical interest in using somatostatin analogs labeled with Auger and conversion electron emitters as antitumor agents. We believe it is crucial to determine whether these compounds or their radiolabeled metabolites gain access to the nucleus. However, we find several inconsistencies in the data and have significant reservations regarding the authors’ conclusion.
The authors suggested that radiolabeled somatostatin analogs were delivered intact to the nucleus. This conclusion is difficult to reconcile with other studies indicating that 111In-pentetreotide, 125I-[Tyr11]-somatostatin, and other somatostatin analogs are degraded after internalization. Viguerie et al. (2) previously showed that 125I-[Tyr11]-somatostatin was rapidly degraded by pancreatic acini, and the radiolabeled metabolite was thought to represent monoiodotyrosine. On the basis of a series of subcellular fractionation and metabolism studies, we previously proposed that 111In-pentetreotide was delivered to tumor cell lysosomes in a tumor-bearing rat model and degraded (3). Two separate studies also determined that 111In-pentetreotide is metabolized to 111In-diethylenetriaminepentaacetic acid-d-Phe in vivo (4,5). We are uncertain why these data were not addressed by Hornick et al. (1). Although it is possible that different cell types may metabolize radiolabeled somatostatin analogs differently or may even deliver them to different compartments, we doubt that this is the explanation for the discrepant results.
Nuclei and lysosomes are dense intracellular compartments, and they often copurify during subcellular fractionation. It is crucial to realize that the fractionation scheme used by Hornick et al. (1) did not separate lysosomes from nuclei, because Figure 1 shows that both DNA and N-acetyl-β-glucosaminidase are concentrated in fractions 12–15. Thus, the subcellular fractionation data (Figs. 5 and 7) with 111In-pentetreotide could simply represent accumulation of the radiolabel within lysosomes. The data in Figure 9 concerning copurification of DNA and radiolabeled somatostatin analogs lack key controls. If 111In-pentetreotide does bind specifically to DNA, it should be straightforward to prove that the binding is specific and saturable. It is crucial to recognize that copurification of DNA with a radiolabeled somatostatin analog does not prove colocalization in the nucleus or specific binding as a receptor–ligand pair.
The subcellular fractionation data with 125I-WOC 4a shown in Figure 4 suffer from the limitations discussed above. However, the finding that 65% of the radiolabel sedimented with nuclei under conditions where >90% of lysosomal enzymes were recovered in the supernatant deserves further consideration because this is the best evidence that the radiolabel was delivered to the nucleus. Important controls would include fractionation after cell surface binding at 4°C. It would also be important to determine the kinetics of nuclear translocation at 37°C and whether this translocation could be inhibited by agents that disassemble the cytoskeleton. Another concern about the experiment is whether the DNA extracting process destroyed the lysosomal membrane, because the preparation kit for genomic tissue DNA isolation contains surfactant. As a result, the radiolabels in the lysosomes may have been released and then may have migrated to the DNA fractions. No control experiments were reported that confirmed that the lysosomes remained intact during the DNA extraction process.
Additional questions include the following:
If 44% of the 111In-pentetreotide counts can still be stripped from the cell surface by acid washing at 24 h (data from Table 1, 20,220 cpm stripped/45,569 total cpm), why is there not a corresponding radioactivity peak that copurifies with the plasma membrane marker 5′nucleotidase in Figure 4C?
Figure 1 shows that the DNA is concentrated in fractions 13 and 14; however, a 24-h incubation in tumor cells with either 125I-WOC 4a or 111In-pentetreotide (Figs. 4C and 5C) shows that the radioactivity is concentrated in fractions that contain less dense subcellular material (fractions 12 and 13 for 125I-WOC 4a and fractions 11 and 12 for 111In-pentetreotide). This discrepancy persists when the marker beads (triangular symbols) are used as internal controls. In Figure 1, DNA copurifies with the heaviest marker beads, but in Figure 4C, this internal standard is found in fraction 15. In Figure 5C, this marker is found in fraction 12.
Why should 111In-pentetreotide distribute nearly uniformly in a Percoll (Sigma Chemical Co., St. Louis, MO) gradient (Fig. 8)? Soluble proteins and peptides should be found at the very top of the gradient.
In summary, the authors should be commended for carrying out research that addresses an important question for the success of the future development of targeted radiotherapy agents. However, we would like to raise awareness of alternative interpretations of their data.
REFERENCES
REPLY:
We appreciate the careful critique of our recently published article (1) by Duncan et al. and would like to respond to their insightful comments. They suggested that our results showing intact radiolabeled somatostatin analogs in the nucleus of cells are difficult to reconcile with those of Viguerie et al. (2), who showed that 125I-[Tyr11]-somatostatin was rapidly degraded by pancreatic acini. Viguerie et al. did show degradation of a labeled somatostatin with a short biologic half-life to iodotyrosine “at the plasma membrane level,” whereas “intracellular and membrane-bound radioactivity was mainly intact 125I-[Tyr11]-somatostatin.” However, these same authors found that 96.3% of the label in the cell interior was intact and that a substantial unmetabolized portion was localized in a nuclear fraction. This result is similar to our findings. However, Viguerie et al. did not specifically isolate lysosomes, although the centrifugal spin used to pellet their nuclear fraction (1500g × 12 min) would not be expected to effectively pellet lysosomes from the cell homogenate. In response to the observation that “rapid intracellular degradation of iodinated proteins and peptides…has been shown by other groups,” we would like to point out that a substantial body of research has also shown that numerous radiolabeled peptides and growth factors, including insulin, prolactin, growth hormone, epidermal growth factor, fibroblast growth factor 2, and nerve growth factor are internalized by endocytosis, translocated to the cytoplasm, and accumulate in the nucleus often bound to chromatin (3,4).
We agree that it is important to realize that our fractionation scheme does not separate lysosomes from nuclei. This is why we refer to the peak in the dense portion of the gradient as “nuclear lysosomal.” This is also why we performed the experiment shown in Figure 6, contrasting the distribution of radioactivity in the gradients with and without nuclei present. These experiments showed that two thirds of the label in the nuclear-lysosomal peak are concentrated in the nuclei.
It is also true, as Duncan et al. suggested, that the DNA extraction procedure would likely lyse the lysosomal membrane, possibly resulting in the release of the radiolabel. We did not control for this possibility, other than by performing the short-term experiment in which the radiolabel was added to the cell homogenate before gradient centrifugation (Fig. 9). This experiment showed no binding to the nuclear fraction.
With respect to questions about the DNA peak and internal consistency, we believe that an internal standard such as density marker beads is essential when comparing gradients with each other. In our study, DNA in Figure 1 is concentrated in the fractions corresponding to the heaviest marker beads and in the one fraction beyond that. In Figures 5B and C and 6, the peaks that show the presence of label in the nucleus correspond exactly to the heaviest marker beads, whereas in Figure 7, the highest peak is in the fraction just beyond that. In all these cases, the highest density marker bead appears in either fraction 12 or fraction 13, whereas inFigure 4C, although our dense peak again comes out in fractions 12 and 13, the highest density marker appears in fraction 15. We believe that the density beads were too dry in this experiment because the density profile does not resemble any of the other experiments. Furthermore, the results of Duncan et al. are sorted by fraction number rather than by density, making comparisons of the two techniques unreliable.
Finally, Duncan et al. compared our work with theirs (5) and suggested that the predominant target for the intracellular localization of the 111In-diethylenetriamine pentaacetic acid–octreotide (111In-DTPA-octreotide) is the lysosome. In contrast to our studies, these authors injected 111In-DTPA-octreotide into tumor-bearing animals and harvested liver, kidney, pancreas, and pancreatic tumors, both 1 h and 20 h after injection. This protocol is very different from our studies, in which human neuroblastoma cells were cultured in a medium that provided a continuous exposure to radioligand for protracted periods. At the cellular level, the difference between these studies is a comparison of a bolus injection with a constant infusion. Our system provides for a protracted exposure of cells to a high environmental concentration of radioligand, exposing cells to a constant receptor-dependent “pressure” to internalize. It can easily be seen that the internal routing of ligand may well be different under these two experimental conditions. Clearly, these two experimental systems cannot be directly compared.
In their studies, Duncan et al. noted that they studied lysosomes in their gradients but made no effort to study the translocation of ligand to a potential nuclear target. However, with respect to our study, these authors have brought up many worthwhile questions. It is unknown what fraction of the nuclear radioligand remains intact after prolonged exposure of somatostatin receptor subtype 2 (sst-2) expressing cells to constant levels of radioligand. Preliminary studies from our laboratory indicate that some of the radioactivity that is progressively translocated to the nucleus is not intact peptide; however, these same studies indicate that up to 30% of the nuclear-associated radioligand is intact 111In-DTPA-octreotide. Clearly, the continuous exposure of receptor-bearing cells to a peptide translocates unmetabolized peptide into the cell in a continuous fashion. In conclusion, it appears that prolonged exposure of sst-2–expressing cells to a constant level of peptide promotes internalization and may provide a significant benefit for the cytotoxic effect of Auger-emitting radiopeptides regardless of their ultimate intracellular destination.
In this issue
Jump to section
Related Articles
Cited By...
- No citing articles found.