Increasing Melanoma Cell Death Using Inhibitors of Protein Disulfide Isomerases to Abrogate Survival Responses to Endoplasmic Reticulum Stress

Penny E. Lovat,1 Marco Corazzari,3 Jane L. Armstrong,2 Shaun Martin,1 Vittoria Pagliarini,3

David Hill,1 Anna M. Brown,1 Mauro Piacentini,3,4 Mark A. Birch-Machin,1

and Christopher P.F. Redfern2

1Dermatological Sciences, School of Clinical and Laboratory Sciences and 2Northern Institute for Cancer Research, Newcastle University,

Newcastle upon Tyne, United Kingdom; 3L’Spallanzani, National Institute for Infectious Disease; and 4University of Tor Vergata, Rome, Italy

Summary:

Malignant melanoma is one of the most difficult cancers to treat due largely in part to its wide range of molecular defenses against immune and cytotoxic attack. It is therefore fitting that much research in chemotherapy drugs has been dedicated toward determining methods to disrupt these molecular defenses. One of the newer forms of chemotherapy acts by inducing endoplasmic reticulum (ER) stress, or disruption of normal ER function, in melanoma cells. Often, homeostatic responses, through proteins such as ER chaperones, can repair the cell and restore its normal function. However, when that homeostatic response is insufficient, the cell will undergo apoptosis.

In the present study, researchers looked to enhance the effect of such chemotherapy drugs (i.e., apoptosis of melanoma cells) by disrupting the homeostatic response to ER stress – specifically, the protein disulfide isomerase (PDI) domains of ER chaperone proteins. In the end, the researchers found that bacitracin, a PDI inhibitor, enhanced the apoptotic effect of chemotherapy drugs fenretinide and velcade.

The first step researchers took in reaching their conclusion was to confirm that the chemotherapy drugs fenretinide and velcade in fact caused ER stress. They exposed three melanoma cell lines (CHL-1, A375 and WM266-4) to various time-courses of each of the two chemotherapy drugs and measured levels of certain ER stress response proteins which are up-regulated during ER stress. Quantitative PCR was used to specifically probe mRNA levels of GADD34, a cellular stress response protein that mediates apoptosis, and ATF4, a stress response transcription factor. Also, DNA electrophoresis was used in tandem with PCR to monitor the mRNA splicing of transcription factor XBP-1, a protein up-regulated during ER stress. The results show that, relative to the control, cells exposed to the chemotherapy drugs had higher levels of GADD34 and ATF4 as well as increased XBP-1 mRNA splicing. This suggested that ER stress was induced by the chemotherapy drugs.

Once ER stress was confirmed, researchers showed that chemotherapy-induced apoptosis was enhanced by the presence of PDI inhibitors, particularly bacitracin, which inhibit the function of chaperone proteins that protect melanoma cells from ER stress. They used flow cytometry of propidium iodide-stained cells to estimate the level of apoptosis and used the MTS assay to measure cell viability of melanoma cells exposed to the following conditions: vehicle control (negative control), 500 μM bacitracin, 10 μM fenretinide, bacitracin + fenretinide, 30 nm velcade, bacitracin + velcade. The results of these experiments are shown in Figure 2B.

In summary, the bacitracin alone did not affect cell viability, but did enhance apoptosis of fenretinide (or velcade) when mixed with that chemotherapy drug. The researchers also used a western blot to show the presence of GADD153, a protein whose increased expression is widely accepted to be a marker of apoptosis due to ER stress. Figure 2C shows that the lanes with both bacitracin and one of the chemotherapy drugs had darker bands in GADD153 for all three melanoma cell lines. It is important to note that the researchers also found that bacitracin does not enhance the effect of chemotherapy drugs on healthy melanocyte cells. This differential effect has numerous implications in chemotherapy drug development.

Finally, to confirm that bacitracin was in fact enhancing apoptosis through the inhibition of PDI activity, the researchers designed a clever controlled experiment. By overexpressing chaperone proteins with PDI domains through transient transfection, the researchers limited the ability of bacitracin to enhance ER stress-induced apoptosis by having the transfected PDI-active proteins compete with endogenous PDI proteins – specifically P4HB – for the inhibition site of bacitracin. By transfecting both wild-type P4HB (with normal PDI activity) and mutant P4HB (with mutations in the PDI active site) and comparing their response to bacitracin, the researcher were able to check to see if bacitracin was in fact acting by inhibiting PDI activity. The results are summarized in Figure 4.

Figure 4B shows that the wild-type PDI cells had higher PDI activity than the mutant variety. This was measured as increased turbidity at 650 nm when reacted with bovine insulin. Figure 4C shows that bacitracin had a lesser effect on enhancing apoptosis in cells with overexpressed wild-type P4HB than cells with overexpressed mutant P4HB. This confirms that bacitracin acts by inhibiting the PDI domains.

Importance/Significance:

This study has invaluable implications for new chemotherapy drug development. Researchers found that PDI inhibiting chemicals, such as bacitracin, can enhance the effect of existing chemotherapy agents. Additionally, these PDI inhibiting chemicals do not increase healthy melanocyte apoptosis in response to chemotherapy drugs, allowing drug developers to enhance tumor cell death without increasing non-tumor cell apoptosis.

Critique:

The authors do a great job planning out their experiments with the appropriate controls and accounting for confounding factors by running additional tests. For example, for each experiment where they expose the melanoma cells to chemotherapy drugs, there is always a set of cells that act as a negative control and do not receive the drugs. Also, when they were unsure of whether bacitracin acted by inhibiting the PDI domains, researchers performed an additional controlled experiment to confirm their suspicions.

On the other hand, the authors did a poor job explaining how they determined the brightness of bands in their western blots. In some cases it was clear which band was darker, but in most cases it was unclear and the paper did not establish a method to quantify darkness of bands. They could have applied a simple ImageJ protocol to strengthen the assertions they make regarding their western blot results.