AJICAP Second Generation: Improved Chemical Site-Specific Conjugation Technology for Antibody-Drug Conjugate Production

The site-directed chemical conjugation of antibodies remains an area of great interest and active efforts within the antibody-drug conjugate (ADC) community. We previously reported a unique site modification using a class of immunoglobulin-G (IgG) Fc-affinity reagents to establish a versatile, streamlined, and site-selective conjugation of native antibodies to enhance the therapeutic index of the resultant ADCs. This methodology, termed "AJICAP", successfully modified Lys248 of native antibodies to produce site-specific ADC with a wider therapeutic index than the Food and Drug Administration-approved ADC, Kadcyla. However, the long reaction sequences, including the reduction-oxidation (redox) treatment, increased the aggregation level. In this manuscript, we aimed to present an updated Fc-affinity-mediated site-specific conjugation technology named "AJICAP second generation" without redox treatment utilizing a "one-pot" antibody modification reaction. The stability of Fc affinity reagents was improved owing to structural optimization, enabling the production of various ADCs without aggregation. In addition to Lys248 conjugation, Lys288 conjugated ADCs with homogeneous drug-to-antibody ratio of 2 were produced using different Fc affinity peptide reagent possessing a proper spacer linkage. These two conjugation technologies were used to produce over 20 ADCs from several combinations of antibodies and drug linkers. The in vivo profile of Lys248 and Lys288 conjugated ADCs was also compared. Furthermore, nontraditional ADC production, such as antibody-protein conjugates and antibody-oligonucleotide conjugates, were achieved. These results strongly indicate that this Fc affinity conjugation approach is a promising strategy for manufacturing site-specific antibody conjugates without antibody engineering.

Biological Evaluation of Maytansinoid-Based Site-Specific Antibody-Drug Conjugate Produced by Fully Chemical Conjugation Approach: AJICAP®

BACKGROUND

Trastuzumab-emtansine (T-DM1, commercial name: Kadcyla) is well-known antibody-drug conjugate (ADC) and was first approved for human epidermal growth factor receptor 2 (HER2)-positive metastatic breast cancer. This molecular format consisting of trastuzumab and maytansinoid payload (emtansine) is very simple, however, T-DM1 has wide heterogeneity due to non-specific conjugation, lowering its therapeutic index (TI).

METHODS

To overcome this issue during the chemical modification of the random conjugation approach to generate T-DM1, we developed a novel chemical conjugation technology termed "AJICAP®" for modification of antibodies in site-specific manner by IgG Fc-affinity peptide based reagents.

RESULTS

In this study, we compared site-specific maytansinoid-based ADCs synthesized by AJICAP and T-DM1 in rat safety studies. The results indicated an increase in the maximum tolerated dose, demonstrating an expansion of the AJICAP-ADC therapeutic index compared with that of commercially available T-DM1. Gram scale preparation of this AJICAP-ADC and the initial stability study are also described.

CONCLUSIONS

Trastuzumab-AJICAP-maytansinoid produced by this unique chemical conjugation methodology showed higher stability and tolerability than commercially available T-DM1.

Chemical Site-Specific Conjugation Platform to Improve the Pharmacokinetics and Therapeutic Index of Antibody-Drug Conjugates

To overcome a lack of selectivity during the chemical modification of native non-engineered antibodies, we have developed a technology platform termed "AJICAP" for the site-specific chemical conjugation of antibodies through the use of a class of IgG Fc-affinity reagents. To date, a limited number of antibody-drug conjugates (ADCs) have been synthesized via this approach, and no toxicological study was reported. Herein, we describe the compatibility and robustness of AJICAP technology, which enabled the synthesis of a wide variety of ADCs. A stability assessment of a thiol-modified antibody synthesized by AJICAP technology indicated no appreciable increase in aggregation or decomposition upon prolonged storage, indicating that the unexpectedly stable thiol intermediate has a great potential intermediate for payload or linker screening or large-scale manufacturing. Payload conjugation with this stable thiol intermediate generated several AJICAP-ADCs. In vivo xenograft studies indicated that the AJICAP-ADCs displayed significant tumor inhibition comparable to benchmark ADC Kadcyla. Furthermore, a rat pharmacokinetic analysis and toxicology study indicated an increase in the maximum tolerated dose, demonstrating an expansion of the AJICAP-ADC therapeutic index, compared with stochastic conjugation technology. This is the first report of the therapeutic index estimation of site-specific ADCs produced by utilizing Fc affinity reagent conjugation. The described site-specific conjugation technology is a powerful platform to enable next-generation ADCs through reduced heterogeneity and enhanced therapeutic index.

Glutamate metabolism in a human intestinal epithelial cell layer model

Plasma glutamate concentrations are constant despite dynamic changes in diets. Most likely, virtually all the dietary glutamate is metabolized in the gut. The present study investigated permeability and metabolism of dietary glutamate in a Caco-2 intestinal epithelial cell layer model by tracing the fate of [U-¹³C] or [¹⁵N]glutamate added to the apical medium. For comparison, several other labelled essential and non-essential amino acids were tested as well. Almost all the labelled glutamate in the apical medium (98% and 96% at 24 h of the culture, respectively) was incorporated in the cell layer, while it barely appeared at the basolateral side, indicating an almost complete utilization of glutamate. Indeed, the ¹³C was incorporated into alanine, proline, ornithine, and glutamine, and the ¹⁵N was incorporated into alanine, glutamine, ornithine, proline, branched chain amino acids and also found as ammonia indicative of oxidation. In contrast, substantial apical-to-basolateral transport of amino acids (8-85% of uptake) other than glutamate and aspartate was evident in studies using amino acid tracers labelled with ¹³C, ¹⁵N or D. These results suggest that the intestinal epithelial cell monolayer utilizes dietary glutamate which adds to maintaining glutamate homeostasis in the body.

The influence of interleukin-6 on the growth of human esophageal cancer cell lines

We reported that human esophageal cancer cell lines (ECC) (YES-1, -2, -3, -4, -5, and -6) produced interleukin-6 (IL-6). We, therefore, investigated the growth effects ([3H]thymidine uptake assay and direct cell count) of IL-6 on these ECC. IL-6 receptor (R) and GP-130 mRNA were detected in all the ECC, using reverse transcriptase-polymerase chain reaction (RT-PCR) assay, and IL-6R was detected in one (YES-3) by immunohistochemical staining. IL-6, anti-IL-6 monoclonal antibody (mAb), or anti-IL-6R mAb caused no reproducible enhancement or suppression of [3H]thymidine uptake by all six ECC. Direct cell count also revealed that the growth enhancement or suppression by IL-6, anti-IL-6 mAb, or anti-IL-6R mAb was relatively small. Particularly, there was no significant sensitivity of YES-3 cells, which definitely produce IL-6 and express IL-6R for IL-6, anti-IL-6 mAb, or anti-IL6R mAb. These results suggest that some esophageal cancers may produce IL-6 and express IL-6R. However, no major interactions between IL-6 and the growth of human esophageal cancer cell lines were detected in this study.

[Optimal protocol for 99mTc-tetrofosmin myocardial SPECT imaging with exercise or dipyridamole stress test and the characteristics of bullseye normal file]

To determine optimal protocol for 99mTc-tetrofosmin SPECT imaging, 30 normal volunteers underwent treadmill or dipyridamole stress test with 770 MBq of tetrofosmin injection. Anterior planar images were acquired every 10-15 minutes for calculating heart/liver ratio. SPECT images were also obtained every 30 minutes for generating Bullseye normal files. In analysis of spatial tracer distribution among normal files, myocardium in Bullseye plot was divided in 9 areas, where mean values of % tracer uptake were calculated and compared by analysis of variance (ANOVA). In those 9 areas, mean values of SD (SD mean) were also compared for predicting artificial blackout in the Bullseye plot among 3 standard normal files: dipyridamole-201Tl, dipyridamole-tetrofosmin (60 min post-injection), and treadmill-tetrofosmin (30 min post-injection). In planar image analysis, high initial uptake in the liver and its acceptable clearance were noted in dipyridamole stress test. Adequate heart/liver ratio was accomplished immediately after treadmill, but 40 minutes later after dipyridamole stress test. ANOVA analysis showed no statistical difference in the spatial % tracer uptake among normal files, irrespective of time after acquisition, data collection time, or difference of tracer. However, SD means in inferior and lateral area of treadmill-tetrofosmin file were significantly small compared to that of dipyridamole-201Tl normal file (p < 0.02 by ANOVA). Thus, we conclude that optimal timing for tetrofosmin SPECT imaging is immediately after treadmill, but at least 40 minutes after dipyridamole stress test. In addition, we should also keep in mind that unexpected blackout may appear in inferior and lateral area, when applying our normal files for reference.