• We are sorry, but NCBI web applications do not support your browser and may not function properly. More information
Logo of pnasPNASInfo for AuthorsSubscriptionsAboutThis Article
Proc Natl Acad Sci U S A. Dec 13, 2005; 102(50): 17987–17992.
Published online Dec 1, 2005. doi:  10.1073/pnas.0509035102
PMCID: PMC1298180
Cell Biology

Oxidizing potential of endosomes and lysosomes limits intracellular cleavage of disulfide-based antibody–drug conjugates


Antibody–drug conjugate therapy entails targeted killing of cancer cells with cytotoxic compounds covalently linked to tumor-specific antibodies and shows promise in the treatment of several human cancers. Current antibody–drug conjugate designs that incorporate a disulfide linker between the antibody and cytotoxic drug are inspired by indirect evidence suggesting that the redox potential within the endosomal system is reducing. It is presumed that antigen-dependent endocytosis leads to disulfide linker reduction and intracellular release of free drug, but direct demonstration of such a mechanism is lacking. To determine whether the disulfide N-succinimidyl 4-(2-pyridyldithio)pentanoate (SPP) linker would be reduced during endocytic recycling of the anti-HER2 antibody trastuzumab (Herceptin, Genentech), we synthesized a trastuzumab-SPP-Rhodamine red conjugate and developed a linker cleavage assay by using the self-quenching property of this fluorophore. In breast carcinoma SKBr3 cells, no SPP linker cleavage was observed, as detected by fluorescence dequenching upon internalization. By contrast, the conjugate did display fluorescence dequenching when diverted to the lysosomal pathway by geldanamycin, an effect partly due to proteolytic degradation rather than disulfide reduction. To understand why linker reduction was inefficient, we measured redox potentials of endocytic compartments by expressing a redox-sensitive variant of GFP fused to various endocytic proteins. Unexpectedly, we found that recycling endosomes, late endosomes, and lysosomes are not reducing, but oxidizing and comparable with conditions in the endoplasmic reticulum. These results suggest that intracellular reduction is unlikely to account for the potency of disulfide-linked antibody–drug conjugates.

Keywords: disulfide linker, redox potential, endocytosis, HER2, Herceptin

One approach to the treatment of cancer is to specifically target cytotoxic drugs to tumor cells by linking them via a cleavable linker to antibodies that recognize a tumor-restricted antigen. Such linkers include hydrazone linkers, designed to hydrolyze upon internalization into acidic endosomes and lysosomes (13); peptide linkers optimized for cleavage by certain lysosomal proteases (35); and disulfide linkers, thought to be cleaved by the reducing environment within the endocytic pathway (610). In the latter category, the monoclonal antibody C242 against CanAg (a glycotope on the mucin1 (MUC1) colorectal tumor antigen) has shown efficacy against colorectal xenograft models in vivo when disulfide-linked to the ribosomal inhibitor ricin A chain via a 4-succinimdyloxycarbonyl-methyl-α-[2-pyridyldithio]-toluene (SMPT) linker (11) and to the maytansinoid-derived microtubule active drug DM1 via an N-succinimidyl 4-(2-pyridyldithio)pentanoate (SPP) linker (12). The latter conjugate is now humanized and in clinical trials as cantuzumab mertansine (9, 13). An ideal linker should be cleaved only upon internalization of the antibody–drug conjugate into the tumor cell, thereby specifically releasing the drug inside tumor cells rather than into the circulation. The aim is to increase the therapeutic index of the antibody–drug conjugate not only by specifically targeting the tumor but also by decreasing systemic toxicity (14).

A validated antibody target for metastatic breast cancer is HER2 (also known as neu or ErbB2), an oncogenic member of the epidermal growth factor receptor family of tyrosine kinases that is overexpressed in ≈25% of cases and is associated with poor prognosis (15). HER2-driven tumor cell growth is inhibited by trastuzumab (Herceptin, Genentech), a humanized anti-HER2 antibody that shows clinical efficacy as a single agent and more recently as adjuvant therapy (1618). Some HER2-overexpressing tumors are resistant to trastuzumab (19) but might still be sensitive to anti-HER2–drug conjugate therapy. Indeed the anti-HER2 antibody TA.1 showed much greater efficacy in vitro against HER2-overexpressing breast carcinoma SKBr3 cells when conjugated via a disulfide N-succinimidyl 3-(2-pyridyldithio)propionate (SPDP) linker to maytansine than as a naked antibody (8), and similar results have been obtained using trastuzumab conjugated to the maytansinoid derivative DM1 via a disulfide SPP linker (20, 21).

We previously showed in SKBr3 cells that HER2-bound trastuzumab is predominantly surface-localized and dynamic, undergoing endocytosis at a rate of ≈1–2% per min, followed by efficient recycling back to the cell surface (22). This endocytic dynamism appears to influence the entire surface pool of Her2, which rapidly and efficiently down-regulates in response to the ansamycin antibiotic geldanamycin (GA) exclusively via improved degradative sorting within endosomes, without increasing the Her2 endocytosis rate. Because protein degradation is known to be more efficient after disulfide bond reduction (2325), and evidence suggests disulfide bonds are cleaved after endocytosis (2629), it is often assumed that the endocytic pathway is reducing, especially among investigators in the field of antibody–drug conjugate therapy. This raises the prospect that SPP linker reduction within recycling and early endosomes accounts for the efficacy of trastuzumab-SPP-DM1. To address whether this is the case, we have developed an intracellular linker cleavage assay by using trastuzumab conjugated via the SPP linker to the fluorophore Rhodamine red (RR) (trastuzumab-SPP-RR). Despite appreciable flux of HER2-bound trastuzumab through the endocytic recycling pathway, disulfide linker cleavage was not efficient, raising the possibility that the redox potential within recycling and early endosomes may not in fact be reducing. Direct measurement of the redox potential of the endocytic pathway has been hampered by the lack of redox probes that are insensitive to their acidic and/or proteolytic environments. To directly measure redox conditions within the endocytic compartments, we have used a redox-sensitive variant of GFP (roGFP1; ref. 30) fused to the luminal sides of various endocytic marker proteins. Surprisingly, we find that recycling endosomes, late endosomes, and lysosomes are not reducing, but rather oxidizing.


Materials, Cell Lines, and Antibodies. See Supporting Text, which is published as supporting information on the PNAS web site.

Preparation of Trastuzumab-SPP-RR. Trastuzumab equilibrated in 35 mM citrate/150 mM NaCl, pH 6.0, was reacted with SPP at 0.5×, 1.5×, and 10× molar ratios for 4 h at 20°C; purified by size-exclusion chromatography in 10 mM succinate/150 mM NaCl, pH 5.0; and reacted overnight at 20°C with 4×, 4×, and 10× molar ratios of RR thiol to produce trastuzumab-SPP-RR with dye/Ab ratios of 0.5, 1.5, and 4.5, respectively, before exchanging into PBS (Fig. 6, which is published as supporting information on the PNAS web site).

Construction and Expression of roGFP1 Constructs. See Supporting Text.

Spectrofluorimetry. Fluorescent constructs were diluted to 20 μg/ml with PBS and measured in an SLM 8000C FP-807 Spectrofluorimeter (SLM Instruments, Urbana, IL) equipped with a double-grating excitation monochromator and a single-grating emission monochromator (RR excitation/emission at 570/588 nm and Alexa-488 excitation/emission at 488/506 nm). Ten millimolar sodium 2-mercaptoethanesulfonate or 0.1 mg/ml subtilisin was then added and fluorescence changes monitored after the indicated incubation times at 22°C with continuous stirring. Data were normalized to the value of the T = 0 time point.

Flow Cytometry. RR mean fluorescence intensity of 10,000 cells gated for exclusion of propidium iodide was quantified with a Coulter Epics Elite-ESP flow cytometer equipped with an Innova 302 krypton ion laser tuned to 568 nm (Coherent, Santa Clara, CA). Alexa-488 mean fluorescence intensity of 10,000 cells gated for exclusion of propidium iodide was quantified with a Beckman Coulter Epics XL-MCL single argon 488-nm laser flow cytometer.

Fluorescence Microscopy. See Supporting Text.

Calculation of Redox Potentials. Calculations were performed essentially as described (30, 31); see Supporting Text for specific details.


A Trastuzumab-SPP-RR Conjugate Displays Fluorescence Self-Quenching That Can Be Used to Report Linker Cleavage. In an effort to visualize and quantify trastuzumab-SPP linker reduction in live cells, we prepared a fluorescent trastuzumab-SPP-RR conjugate (Fig. 6). The SPP-RR moiety was conjugated to free amines (on lysine residues) on trastuzumab at three different labeling ratios of 0.5, 1.5, and 4.5 fluorophores per antibody (trastuzumab-SPP-RR0.5, trastuzumab-SPP-RR1.5, and trastuzumab-SPP-RR4.5, respectively). The excitation and emission spectra of RR partially overlap, which may explain in part the fluorescence self-quenching observed when xanthene-type dyes are crowded close together (32, 33). Indeed, when the dimerized disulfide form of the synthetic intermediate RR thiol was reduced back to the monomeric state in vitro, the fluorescence intensity increased >2-fold (Fig. 1A, diamonds). A similar fluorescence dequenching effect was observed upon in vitro reduction of the highly labeled conjugate, trastuzumab-SPP-RR4.5 (Fig. 1A, squares). Fluorescence dequenching was much less pronounced for trastuzumab-SPP-RR1.5 and completely absent for trastuzumab-SPP-RR0.5 (Fig. 1A, triangles and circles, respectively), consistent with the idea that self quenching requires close proximity of two or more RR moieties. In support of this explanation, in vitro proteolytic digestion of trastuzumab-SPP-RR4.5 also resulted in 2-fold fluorescence dequenching (Fig. 1B, squares). In contrast, Alexa-488-conjugated trastuzumab showed very little dequenching upon proteolytic digestion (Fig. 1B, triangles), suggesting that intrinsic fluorophore properties, more pronounced fluorophore crowding, or both rendered the trastuzumab-SPP-RR4.5 construct more susceptible to self quenching. These data indicate that fluorescence dequenching of trastuzumab-SPP-RR4.5 can report SPP linker cleavage, although it may also report other events that relieve fluorophore crowding, such as proteolytic fragmentation of the antibody.

Fig. 1.
Fluorescence dequenching of a trastuzumab-SPP-RR conjugate upon SPP linker cleavage in vitro.(A) Crowding of RR moieties, as in the oxidized disulfide dimer form of RR thiol (diamonds) or trastuzumab-SPP-RR conjugate with 4.5 RR/antibody (squares), attenuates ...

Trastuzumab-SPP-RR Linker Cleavage Is Inefficient in Breast Carcinoma SKBr3 Cells, Despite Continuous Endocytic Recycling. If the SPP linker is reduced within the endocytic pathway as presumed (12), it may be detectable as fluorescence dequenching of prebound trastuzumab-SPP-RR4.5 upon internalization. In human breast carcinoma SKBr3 cells, the trastuzumab endocytosis rate is as high as ≈1.5% per minute, and both the antibody and antigen are efficiently recycled back to the plasma membrane (22). Thus the surface-bound pool of trastuzumab appears to be subjected to considerable flux through the endocytic recycling pathway, corresponding to a first-order “endocytic exposure” half-life of ≈45 min. Despite this significant exposure to endosomes, fluorescence dequenching of surface-bound trastuzumab-SPP-RR4.5 was not efficient in SKBr3 cells after up to 150-min incubation at 37°C [Fig. 2, control (circles)], increasing < 20%. By contrast, a >2-fold dequenching effect was observed in the presence of the ansamycin antibiotic GA (Fig. 2, diamonds), which we previously showed diverts HER2 from the early endosomes to the lysosomal degradative pathway (22). Conjugate proteolysis within the lysosomal pathway accounted in part for this dequenching effect, which was inhibited nearly 40% by either lysosomal protease inhibitors (PIs) or lysosomal deacidification with chloroquine (34) (Fig. 2, triangles and squares, respectively). Fluorescence dequenching required conjugate internalization, because it was mostly prevented by inhibition of clathrin-mediated endocytosis by hypertonic shock (Fig. 2, crosses), a treatment that inhibited uptake of transferrin and trastuzumab in the presence of GA by ≈90% (Fig. 7, which is published as supporting information on the PNAS web site). These results were corroborated by fluorescence microscopy, which revealed a GA-induced increase in red fluorescence intensity associated with the distribution shift from plasma membrane to endosomes (Fig. 3). In these experiments, we used a dual-labeled conjugate in which the green fluorophore Alexa-488 was linked to free amines of trastuzumab-SPP-RR4.5 directly (without an intervening SPP linker) to serve as an internal intensity reference. In this way, pixel intensity changes attributed intrinsically to RR dequenching, which affects the red channel exclusively, could be distinguished from pixel intensity changes arising from conjugate redistribution and concentration within internal compartments, which affects both red and green channels equally. In support of the flow cytometry results above, GA treatment redistributed the 488trastuzumab-SPP-RR4.5 conjugate from a predominantly surface-localized pattern in which the merged fluorescence signal was mostly green (Fig. 3C) to a predominantly late endosome pattern (see ref. 22 for colocalization data) in which the merged fluorescence signal was yellow (Fig. 3F), indicating that the red signal had intensified relative to the green. Thus, GA induced dequenching of trastuzumab-SPP-RR4.5 via endocytic routing to the lysosomal degradation pathway. Most significantly, the lack of dequenching observed in the absence of GA suggests that exposure to endocytic recycling over this period does not lead to significant intracellular cleavage of the SPP linker.

Fig. 2.
Fluorescence dequenching of trastuzumab-SPP-RR4.5 in live SKBr3 cells requires endocytosis and lysosomal routing. SKBr3 cells with surface-bound trastuzumab-SPP-RR4.5 were treated and incubated at 37°C for the indicated times and fluorescence ...
Fig. 3.
Red fluorescence dequenching of a dual labeled 488trastuzumab-SPP-RR4.5 conjugate upon GA-induced routing to the lysosomal pathway. The green fluorophore Alexa-488 was conjugated to trastuzumab-SPP-RR4.5 to produce a dual-labeled 488trastuzumab-SPP-RR ...

Endosomes and Lysosomes Are Oxidizing Rather than Reducing. The above results call into question the idea that the redox potential within endosomes and lysosomes is reducing. To directly address this issue, we used an engineered roGFP1, which has two cysteine residues on either side of its chromophore (S147C and Q204C) and is pH-insensitive, lacking the S65T substitution (30). The fluorescence excitation maxima at 400 and 490 nm vary ratiometrically with the presence or absence of a disulfide bond between these cysteines, such that reduced roGFP1 has a lower 400-nm peak and a higher 490-nm peak compared with oxidized roGFP1. The 400/490-nm ratio thus reports the redox state of the compartment in which it is expressed (30, 31). We therefore fused roGFP1 to the luminal side of various endocytic compartment marker proteins (35): transferrin receptor to label the recycling endosomes, lysosomal-associated membrane protein (Lamp2a) for the limiting membrane of late endosomes and lysosomes, and CD63 for the internal membranes of lysosomes (but facing the same lysosomal lumen as Lamp2a). As an oxidative compartment control, we fused roGFP1 to calnexin (Cnx), an endoplasmic reticulum (ER) resident protein; and as a reducing compartment control, we expressed mito-ss-roGFP1, in which roGFP has a mitochondrial leader sequence and is imported into the mitochondrial matrix (30, 31). Because SKBr3 cells proved difficult to transfect, PC3 cells were used primarily for this analysis. To avoid overexpression and mislocalization of endosomal constructs to the ER, moderately expressing stable lines were generated. All roGFP fusion proteins were correctly localized, as shown in Fig. 4: mito-ss-roGFP1 fluoresced within the mitochondrial matrix demarcated by cytochrome c (Fig. 4 A–C), roGFP1-Cnx showed good overlap with the ER marker protein disulfide isomerase (Fig. 4 D–F), transferrin receptor-roGFP1 overlapped with Alexa555-transferrin uptaken for 1 h (Fig. 4 G–I), and roGFP1-CD63 (Fig. 4 J–L) and roGFP1-Lamp2a (Fig. 4 M–O) colocalized well with the late endosomal/lysosomal Lamp1.

Fig. 4.
roGFP1-tagged compartment proteins localize correctly in stably transfected PC3 cells. PC3 cells stably transfected with various roGFP1-tagged compartment proteins (A, D, G, J, and M, green channel) were fixed with 3% paraformaldehyde, permeabilized with ...

We then measured roGFP1 fluorescence image pixel intensities at 380-versus 490-nm excitation to calculate the 380/490-nm ratio for each fusion construct (Fig. 5 and Table 1). Measurements after cell treatment with membrane-permeable oxidants (aldithriol or H2O2) or reductants (DTT) were used to determine the 380/ 490-nm ratio of fully oxidized and fully reduced roGFP1 and allowed determination of the percentage of roGFP1 as described (30). As expected, mitochondria-localized mito-ss-roGFP1 was only 8.3% ± 5.4% oxidized, and roGFP1-Cnx was almost completely (95.8 ± 9.7%) oxidized (Fig. 5A and Table 1). Surprisingly, however, the recycling endosomal transferrin receptor and the late endosomal/lysosomal markers Lamp2a and CD63 all showed high roGFP1 380/490 ratios (≈94–97% oxidized) comparable to that of the ER marker Cnx, indicating that in PC3 cells these compartments are oxidizing. Although precise determination of redox potential from the roGFP1 titration curve requires oxidation percentages between ≈10% and 90% (see figure 1C of ref. 30) showing the midpoint potential of roGFP is –288 mV), we estimate that the endocytic pathway compartments are at least as oxidizing as –240 mV, compared with a much more reducing value of –318 mV for the mitochondria.

Fig. 5.
roGFP1 fluorescence intensity measurements reveal that the endocytic pathway is oxidizing. (A) PC3 cells stably transfected with mito-ss-roGFP1 (Mito), roGFP1-calnexin (Cnx), transferrin receptor-roGFP1 (TrfR), roGFP1-Lamp2a (Lamp), and roGFP1-CD63 (CD63) ...
Table 1.
380/490 fluorescence intensity ratios and calculated percent oxidized (all ± SD) for each subcellular compartment

Although an oxidizing potential of the recycling endosomes accounts for the lack of trastuzumab-SPP-RR cleavage in the above experiments, the predominant view that the endocytic pathway is reducing (2325, 28, 29, 36) prompted us to verify that our oxidizing measurements were not artifacts. roGFP1-CD63 was used for all of the following experiments, because it emitted the most robust fluorescence signal. First, the 380/490-nm ratio was unchanged after overnight incubation with PIs, leupeptin, and pepstatin, suggesting that oxidizing measurements were not an artifact of partial proteolysis (Fig. 5B and Table 1, compare “control” with “+PIs”). Second, elimination of newly synthesized proteins upon a 16 h chase with the protein synthesis inhibitor cycloheximide (Fig. 8, which is published as supporting information on the PNAS web site) yielded an unchanged oxidizing ratio, ruling out nascent roGFP1 in the ER as an artifactual source of the oxidized signal (Fig. 5B and Table 1, compare “control” with “+Chx”). Third, cells cultured in 3% rather than 21% oxygen to more closely replicate redox conditions that normally exist in tissues in vivo (37) showed unchanged roGFP1 oxidation at 90.5 ± 7.5% oxidized (Fig. 5B and Table 1, compare “control” with “low O2”). In these experiments, fluorescence measurements were taken live in the presence of the oxygen-depleting enzyme Oxyrase, and we observed a slight but uniform drop in all 380/490-nm ratio measurements, including those in the presence of oxidizing or reducing agents.

To verify that these observations in PC3 cells were also applicable to SKBr3 cells, we retrovirally generated a stable population of SKBr3 cells with low-level expression of roGFP1-CD63 to ensure exclusive localization to late endosomes/lysosomes. The dim signal resulted in a smaller dynamic range and higher signal-to-noise ratio than in PC3 cells and unfortunately precluded similar analysis with the transferrin receptor. Nevertheless, as observed in PC3 cells, roGFP-CD63 was highly oxidized in SKBr3 cells (Table 1), suggesting that the oxidizing potential of the endosomes and lysosomes is not a peculiarity of PC3 cells.


We have developed an assay for monitoring SPP linker cleavage in HER2-overexpressing SKBr3 cells by measuring fluorescence dequenching of trastuzumab-SPP-RR4.5. Despite rapid and continuous cycling of surface-bound trastuzumab-SPP-RR4.5 through endosomes, the SPP linker was not efficiently reduced. Although lysosomal routing in response to GA did result in dequenching, thereby providing a positive control for the assay, results with lysosomal inhibitors indicated that proteolytic degradation rather than linker reduction accounted for at least part of this effect. Because it is unlikely that these inhibitors completely prevented proteolytic cleavage within lysosomes, proteolysis is likely the predominant mechanism of dequenching. Indeed, we observed that endosomes and lysosomes were oxidizing, as assessed by the virtually complete oxidation of roGFP1 fusion proteins expressed in those compartments. The redox values of these compartments were comparable to that of the ER, a compartment with well described oxidative properties (3841). By contrast, and in agreement with previously published results using the same roGFP1 system (30), we found that mitochondria had a reductive potential.

The finding that the endocytic pathway is oxidizing was initially surprising given the assumption that endosomes and/or lysosomes are reducing. A reductive milieu in lysosomes would be expected to facilitate degradation, because proteolysis in vitro is more efficient after protein reduction (28, 42, 43). Indeed, intracellular antigen processing by antigen-presenting cells is more efficient if the antigen is reduced and alkylated before uptake (28, 43). A mechanism for the maintenance of lysosome reductive potential was proposed after the discovery of a rapid lysosomal importer specific for cysteine and possibly cysteamine, both of which have a free thiol group able to mediate reduction of disulfide bonds (36). It has not yet been determined whether this lysosomal cysteine importer is localized on all lysosomes, or just a subset, such as terminal or dense core lysosomes. Because we measured the average fluorescent intensity of all of the given endocytic compartments in all of the cells in any one field together, we cannot exclude the possibility that a minor subset of lysosomes could be reducing, but that these were masked by the oxidizing values of the majority of the earlier endolysosomes. In addition, we cannot exclude the possibility that within any one endosome or lysosome, there are subregions of reducing potential. Although we did not detect any difference in the oxidizing environment of two different lysosomal membrane proteins facing the lysosomal lumen from the outer (Lamp2a) and inner (CD63) membranes, we did not assess the middle of the lumen by using a soluble lysosomal protein, for example.

Previous observations of others, however, support our observations that endosomes and lysosomes are oxidizing. Collins et al. (28) demonstrated that reduction of disulfide-linked transferrin-[125I]-tyrosine conjugates in T cells is at best very inefficient, with a [125I]-tyrosine release rate of ≈6% per hour. This agrees with our inefficient reduction of trastuzumab-SPP-RR, since trastuzumab recycles with similar kinetics to transferrin in SKBr3 cells (22). Furthermore, just as we observed increased dequenching upon GA-induced diversion of trastuzumab-SPP-RR4.5 from the recycling to the lysosomal pathway, these investigators demonstrated efficient release (≈1% per min) of [125I]-tyrosine when it was disulfide-linked instead to the lysosomally targeted α2-macroglobulin. Consistent with our finding that lysosomes are oxidizing, Feener et al. (44) used subcellular fractionation to argue that the disulfide-linked conjugate [125]I-tyramine-SPDP-poly(d-lysine) was not reduced upon delivery to lysosomes. Although our assay may be less sensitive and quantitative than the above radioactive assays, it does offer the ability to image compartments within which reduction (and/or degradation) occurs in intact living cells, without any of the potential spatial and temporal artifacts associated with subcellular fractionation.

If, as our results suggest, lysosomes are oxidizing, why does this not impede the proteolytic functions of this organelle? Considering the complexity of the lysosomal degradation machinery (45), it may be that proteolytic activity within this compartment is so efficient that reduction of disulfides is not necessary. Alternatively, protein disulfides may be rendered labile within lysosomes despite the oxidative conditions, by analogy to how protein disulfide isomerase within the oxidative ER renders the disulfides of nascent secretory proteins labile to facilitate proper protein folding (46). Whereas disulfide lability within the ER encourages proper nascent protein folding mediated by resident chaperones like BiP [binding protein; (47)], such lability within the acidic and degradative lysosomes would facilitate protein unfolding and proteolysis. Indeed, in antigen-presenting cells, IFN-γ-inducible lysosomal thiol reductase (GILT) appears to support processing of at least some antigens (4851). It remains to be determined, however, whether GILT or other oxido-reductases within lysosomes recognize the disulfide SPP or SPDP linkers within antibody–drug conjugates as substrates. Our data suggest that trastuzumab-SPP-RR linker reduction within lysosomes is at best not much more efficient than antibody degradation.

If these disulfide linkers are inefficiently reduced within the endocytic pathway, then how can we account for the in vitro and in vivo potency of disulfide-linked antibody–DM1 conjugates such as cantuzumab mertansine [C242-SPP-DM1 (12)], anti-PSCA-DM1 (52), anti-CD56 huN901-SPP-DM1 (53), anti-HER2-SPDP-DM1 (8), and trastuzmab-SPP-DM1 (20, 21)? Considering that in vitro cytotoxicity studies generally require at least 2–3 days to detect cell proliferation effects, and in vivo studies require even longer to observe effects on tumor growth, it may be that very slow delivery of free DM1, via inefficient SPP linker reduction, inefficient degradation of the recycling antibody, or both, sufficiently accounts for conjugate potency. Although trastuzumab predominantly recycles, it is slowly degraded within SKBr3 cells with a τ1/2 of ≈19 h (22) and is detected readily within lysosomes in cells incubated with PIs for >24 h (C.D.A. and S.J.S., unpublished data). Another possibility is that SPP linker cleavage at the cell surface via plasma membrane reductases (54) or even protein disulfide isomerase itself (55) generates extracellular free DM1 locally in sufficient amounts to kill the cell. Although we found that hypertonic shock inhibited GA-induced dequenching, suggesting that internalization was required for the effect under these conditions, we cannot exclude the possibility that in the absence of GA, some surface SPP cleavage occurs. Our dequenching assay is not well suited to detect surface cleavage, particularly if such cleavage is not efficient. If, however, surface cleavage were very efficient (>50%), it would cause a decrease in fluorescence over time due to release of free RR-thiol from the cell, which was not observed. The results of Feener et al. (44) suggest that in Chinese hamster ovary cells, a small amount (≈3%) of disulfide-linked [125]I-tyramine-SPDP-poly(d-lysine) was reduced at the surface to release free [125]I-tyramine within ≈20 min, because membrane-impermeant sulfhydryl reagents such as DTNB inhibited such release. Such surface reduction events might depend critically on the redox potential of the extracellular environment. We did not attempt to measure the redox potential of the outer plasma membrane, because such a measurement would merely reflect the redox state of the cell culture medium. We are currently attempting to target roGFP to the cell surface of xenograft tumors in mice to measure the redox potential in vivo. In the meantime, it would be informative to determine whether SPP-DM1 conjugated to a non-internalizing antibody would result in tumor regression, because if DM1 were released at the plasma membrane, it should still be able to diffuse into the tumor cells and its neighbors.


Our data show that the endocytic pathway is as oxidizing as the ER, and that the reductive cleavage of the disulfide linker SPP within the recycling endosomal pathway is very inefficient. These results suggest that disulfide-based linkers such as SPP may not be an appropriate or necessary design feature for antibody–drug conjugates.

Supplementary Material

Supporting Information:


We dedicate this paper to the memory of our much-missed colleagues, Susan Palmieri and Ralph Schwall. We are indebted to Dr. Jim Remington (University of Oregon, Eugene) for kindly providing the roGFP and mito-ss-roGFP plasmids and sharing results before publication and Dr. Gary Nolan (Stanford University, Stanford, CA) for the Phoenix cells. We also thank Dr. Susan Palmieri for setup and maintenance of the fluorescence microscope; Bill Forrester for help with the standard deviation calculations; and Drs. Jean-Philippe Stephan, Bill Mallet, Dick Vandlen, Dott Bennett, Vikas Duvurri, and Colette Dooley for helpful discussions.


Author contributions: C.D.A. and S.J.S. designed research; C.D.A. and X.W. performed research; C.D.A., X.W., L.G., C.N., and S.J.S. contributed new reagents/analytic tools; C.D.A., X.W., R.H.S., and S.J.S. analyzed data; and C.D.A., X.W., and S.J.S. wrote the paper.

Conflict of interest statement: All authors are employees of Genentech, Inc.

Abbreviations: Cnx, calnexin; ER, endoplasmic reticulum; GA, geldanamycin; Lamp, lysosmal-associated membrane protein; PIs, lysosomal protease inhibitors; roGFP, redox-sensitive GFP variant; RR, Rhodamine red; SPDP, N-succinimidyl 3-(2-pyridyldithio)propionate (and derivatives); SPP, N-succinimidyl 4-(2-pyridyldithio)pentanoate (and derivatives).


1. Braslawsky, G. R., Kadow, K., Knipe, J., McGoff, K., Edson, M., Kaneko, T. & Greenfield, R. S. (1991) Cancer Immunol. Immunother. 33, 367–374. [PubMed]
2. Hamann, P. R., Hinman, L. M., Beyer, C. F., Lindh, D., Upeslacis, J., Flowers, D. A. & Bernstein, I. (2002) Bioconjug. Chem. 13, 40–46. [PubMed]
3. Doronina, S. O., Toki, B. E., Torgov, M. Y., Mendelsohn, B. A., Cerveny, C. G., Chace, D. F., DeBlanc, R. L., Gearing, R. P., Bovee, T. D., Siegall, C. B., et al. (2003) Nat. Biotechnol. 21, 778–784. [PubMed]
4. Trouet, A., Masquelier, M., Baurain, R. & Deprez-De Campeneere, D. (1982) Proc. Natl. Acad. Sci. USA 79, 626–629. [PMC free article] [PubMed]
5. Jeffrey, S. C., Torgov, M. Y., Andreyka, J. B., Boddington, L., Cerveny, C. G., Denny, W. A., Gordon, K. A., Gustin, D., Haugen, J., Kline, T., et al. (2005) J. Med. Chem. 48, 1344–1358. [PubMed]
6. Scott, C. F., Jr., Goldmacher, V. S., Lambert, J. M., Jackson, J. V. & McIntyre, G. D. (1987) J. Natl. Cancer Inst. 79, 1163–1172. [PubMed]
7. Thorpe, P. E., Wallace, P. M., Knowles, P. P., Relf, M. G., Brown, A. N., Watson, G. J., Blakey, D. C. & Newell, D. R. (1988) Cancer Res. 48, 6396–6403. [PubMed]
8. Chari, R. V., Martell, B. A., Gross, J. L., Cook, S. B., Shah, S. A., Blattler, W. A., McKenzie, S. J. & Goldmacher, V. S. (1992) Cancer Res. 52, 127–131. [PubMed]
9. Smith, S. V. (2004) Curr. Opin. Mol. Ther. 6, 666–674. [PubMed]
10. Henry, M. D., Wen, S., Silva, M. D., Chandra, S., Milton, M. & Worland, P. J. (2004) Cancer Res. 64, 7995–8001. [PubMed]
11. Calvete, J. A., Newell, D. R., Wright, A. F. & Rose, M. S. (1994) Cancer Res. 54, 4684–4690. [PubMed]
12. Liu, C., Tadayoni, B. M., Bourret, L. A., Mattocks, K. M., Derr, S. M., Widdison, W. C., Kedersha, N. L., Ariniello, P. D., Goldmacher, V. S., Lambert, J. M., et al. (1996) Proc. Natl. Acad. Sci. USA 93, 8618–8623. [PMC free article] [PubMed]
13. Tolcher, A. W., Ochoa, L., Hammond, L. A., Patnaik, A., Edwards, T., Takimoto, C., Smith, L., de Bono, J., Schwartz, G., Mays, T., et al. (2003) J. Clin. Oncol. 21, 211–222. [PubMed]
14. Polakis, P. (2005) Curr. Opin. Pharmacol. 5, 382–387. [PubMed]
15. Slamon, D. J., Clark, G. M., Wong, S. G., Levin, W. J., Ullrich, A. & McGuire, W. L. (1987) Science 235, 177–182. [PubMed]
16. Cobleigh, M. A., Vogel, C. L., Tripathy, D., Robert, N. J., Scholl, S., Fehrenbacher, L., Wolter, J. M., Paton, V., Shak, S., Lieberman, G. & Slamon, D. J. (1999) J. Clin. Oncol. 17, 2639–2648. [PubMed]
17. Hynes, N. E. & Lane, H. A. (2005) Nat. Rev. Cancer 5, 341–354. [PubMed]
18. Hanrahan, E. O., Hennessy, B. T. & Valero, V. (2005) Exp. Opin. Pharmacother. 6, 1477–1491. [PubMed]
19. Neve, R. M., Lane, H. A. & Hynes, N. E. (2001) Ann. Oncol. 12, Suppl. 1, S9–S13. [PubMed]
20. Ranson, M. & Sliwkowski, M. X. (2002) Oncology 63, Suppl. 1, 17–24. [PubMed]
21. Erickson, S., Schwall, R. & Sliwkowski, M. X. (2002) U.S. Patent 2002/0001587 A1.
22. Austin, C. D., De Maziere, A. M., Pisacane, P. I., van Dijk, S. M., Eigenbrot, C., Sliwkowski, M. X., Klumperman, J. & Scheller, R. H. (2004) Mol. Biol. Cell 15, 5268–5282. [PMC free article] [PubMed]
23. Bohley, P. & Seglen, P. O. (1992) Experientia 48, 151–157. [PubMed]
24. Jensen, P. E. (1995) Semin. Immunol. 7, 347–353. [PubMed]
25. Pillay, C. S., Elliott, E. & Dennison, C. (2002) Biochem. J. 363, 417–429. [PMC free article] [PubMed]
26. Shen, W. C., Ryser, H. J. & LaManna, L. (1985) J. Biol. Chem. 260, 10905–10908. [PubMed]
27. Papini, E., Rappuoli, R., Murgia, M. & Montecucco, C. (1993) J. Biol. Chem. 268, 1567–1574. [PubMed]
28. Collins, D. S., Unanue, E. R. & Harding, C. V. (1991) J. Immunol. 147, 4054–4059. [PubMed]
29. Fivaz, M., Vilbois, F., Thurnheer, S., Pasquali, C., Abrami, L., Bickel, P. E., Parton, R. G. & van der Goot, F. G. (2002) EMBO J. 21, 3989–4000. [PMC free article] [PubMed]
30. Hanson, G. T., Aggeler, R., Oglesbee, D., Cannon, M., Capaldi, R. A., Tsien, R. Y. & Remington, S. J. (2004) J. Biol. Chem. 279, 13044–13053. [PubMed]
31. Dooley, C. T., Dore, T. M., Hanson, G. T., Jackson, W. C., Remington, S. J. & Tsien, R. Y. (2004) J. Biol. Chem. 279, 22284–22293. [PubMed]
32. MacDonald, R. I. (1990) J. Biol. Chem. 265, 13533–13539. [PubMed]
33. Lakowicz, J. R., Malicka, J., D'Auria, S. & Gryczynski, I. (2003) Anal. Biochem. 320, 13–20. [PMC free article] [PubMed]
34. Helenius, A., Marsh, M. & White, J. (1982) J. Gen. Virol. 58, 47–61. [PubMed]
35. Mellman, I. (1996) Annu. Rev. Cell Dev. Biol. 12, 575–625. [PubMed]
36. Pisoni, R. L., Acker, T. L., Lisowski, K. M., Lemons, R. M. & Thoene, J. G. (1990) J. Cell Biol. 110, 327–335. [PMC free article] [PubMed]
37. Heyneman, C. A. & Lawless-Liday, C. (2002) Crit. Care Nurse 22, 52–60. [PubMed]
38. Hwang, C., Sinskey, A. J. & Lodish, H. F. (1992) Science 257, 1496–1502. [PubMed]
39. Majoul, I., Ferrari, D. & Soling, H. D. (1997) FEBS Lett. 401, 104–108. [PubMed]
40. Kaufman, R. J. (1999) Genes Dev. 13, 1211–1233. [PubMed]
41. Noiva, R. (1999) Semin. Cell Dev. Biol. 10, 481–493. [PubMed]
42. Mego, J. L. (1984) Biochem. J. 218, 775–783. [PMC free article] [PubMed]
43. Jensen, P. E. (1991) J. Exp. Med. 174, 1121–1130. [PMC free article] [PubMed]
44. Feener, E. P., Shen, W. C. & Ryser, H. J. (1990) J. Biol. Chem. 265, 18780–18785. [PubMed]
45. Lockwood, T. D. (2000) Antioxid. Redox Signal. 2, 851–878. [PubMed]
46. Gilbert, H. F. (1997) J. Biol. Chem. 272, 29399–29402. [PubMed]
47. Kassenbrock, C. K., Garcia, P. D., Walter, P. & Kelly, R. B. (1988) Nature 333, 90–93. [PubMed]
48. Arunachalam, B., Phan, U. T., Geuze, H. J. & Cresswell, P. (2000) Proc. Natl. Acad. Sci. USA 97, 745–750. [PMC free article] [PubMed]
49. Phan, U. T., Arunachalam, B. & Cresswell, P. (2000) J. Biol. Chem. 275, 25907–25914. [PubMed]
50. Maric, M., Arunachalam, B., Phan, U. T., Dong, C., Garrett, W. S., Cannon, K. S., Alfonso, C., Karlsson, L., Flavell, R. A. & Cresswell, P. (2001) Science 294, 1361–1365. [PubMed]
51. Sinnathamby, G., Maric, M., Cresswell, P. & Eisenlohr, L. C. (2004) J. Immunol. 172, 6607–6614. [PubMed]
52. Ross, S., Spencer, S. D., Holcomb, I., Tan, C., Hongo, J., Devaux, B., Rangell, L., Keller, G. A., Schow, P., Steeves, R. M., et al. (2002) Cancer Res. 62, 2546–2553. [PubMed]
53. Tassone, P., Gozzini, A., Goldmacher, V., Shammas, M. A., Whiteman, K. R., Carrasco, D. R., Li, C., Allam, C. K., Venuta, S., Anderson, K. C., et al. (2004) Cancer Res. 64, 4629–4636. [PubMed]
54. Varandani, P. T., Raveed, D. & Nafz, M. A. (1978) Biochim. Biophys. Acta 538, 343–353. [PubMed]
55. Mandel, R., Ryser, H. J., Ghani, F., Wu, M. & Peak, D. (1993) Proc. Natl. Acad. Sci. USA 90, 4112–4116. [PMC free article] [PubMed]

Articles from Proceedings of the National Academy of Sciences of the United States of America are provided here courtesy of National Academy of Sciences
PubReader format: click here to try


Related citations in PubMed

See reviews...See all...


  • Compound
    PubChem Compound links
  • PubMed
    PubMed citations for these articles
  • Substance
    PubChem Substance links

Recent Activity

Your browsing activity is empty.

Activity recording is turned off.

Turn recording back on

See more...