Fractionated Radiotherapy with 3 x 8 Gy Induces Systemic Anti-Tumour Responses and Abscopal Tumour Inhibition without Modulating the Humoral Anti-Tumour Response

Accumulating evidence indicates that fractionated radiotherapy (RT) can result in distant non-irradiated (abscopal) tumour regression. Although preclinical studies indicate the importance of T cells in this infrequent phenomenon, these studies do not preclude that other immune mechanisms exhibit an addition role in the abscopal effect. We therefore addressed the question whether in addition to T cell mediated responses also humoral anti-tumour responses are modulated after fractionated RT and whether systemic dendritic cell (DC) stimulation can enhance tumour-specific antibody production. We selected the 67NR mammary carcinoma model since this tumour showed spontaneous antibody production in all tumour-bearing mice. Fractionated RT to the primary tumour was associated with a survival benefit and a delayed growth of a non-irradiated (contralateral) secondary tumour. Notably, fractionated RT did not affect anti-tumour antibody titers and the composition of the immunoglobulin (Ig) isotypes. Likewise, we demonstrated that treatment of tumour-bearing Balb/C mice with DC stimulating growth factor Flt3-L did neither modulate the magnitude nor the composition of the humoral immune response. Finally, we evaluated the immune infiltrate and Ig isotype content of the tumour tissue using flow cytometry and found no differences between treatment groups that were indicative for local antibody production. In conclusion, we demonstrate that the 67NR mammary carcinoma in Balb/C mice is associated with a pre-existing antibody response. And, we show that in tumour-bearing Balb/C mice with abscopal tumour regression such pre-existing antibody responses are not altered upon fractionated RT and/or DC stimulation with Flt3-L. Our research indicates that evaluating the humoral immune response in the setting of abscopal tumour regression is not invariably associated with therapeutic effects.


Development of cellular read-out systems for the detection of TBAs in the plasma of
Balb/C mice.
In order to have a system that allows the detection of tumour-specific antibodies we set out to develop cellular-based detection assays using flow cytometry and ELISA. In order to establish whether these assays quantitatively reflect the antibody concentration or whether they are influenced by undesired protein factors in the plasma sample also defined as matrix As shown in S2a Fig (right panel), the H-2K d binding to the cells was dependent on the plasma dilution. In particular, a plasma dilution of 1:2 showed reduced binding of the H-2K d recognizing antibody to the 67NR tumour cells. Additionally, in this plasma dilution no dose-dependency could be observed when more of the monoclonal antibody was added.
These matrix effects were only observed in the 1:2 dilution of the plasma and all other plasma dilutions demonstrated a dose-dependent increase in antibody binding. Notably, the MFI values were highest in the condition where the plasma was diluted 1:10 and 1:20 and these MFIs exceed the MFI values observed in a typical buffer used for flow cytometry (PBS + 10% FCS + 0.05% Tween 20) with a lower protein content. With exceeding dilution of the plasma the MFI values resemble more closely those that are observed in buffer. Based on these data the highest signal to noise ratio was observed at dilutions above 1:10 and a plasma dilution of 1:20 was used for all antibody detection experiments. Using this serum 2 concentration there is a dose-dependent correlation between the MFI and the quantity of the antibody that is spiked in the plasma.
In order to evaluate whether this flow cytometric methodology also allows the detection of a de novo induced polyclonal antibody response in an animal, we vaccinated healthy Balb/C mice with 10 7 heat killed 67NR cells in incomplete Freund's adjuvant and CpG. Besides the generation of positive control sera, the secondary aim of this experiment was to study whether it is possible to generate antibodies against the 67NR tumour cell line In addition to the development of a flow cytometric detection system we also set out to develop an ELISA-based system for the detection of humoral antibodies. To this end we seeded tissue culture plates with different numbers of 67NR cells and incubated them for 18 hours. Although, 67NR cells showed a strong adherent phenotype (fibroblast-like structure), they detached from the culture plates during these washing steps. Therefore, we fixated 67NR cells with formaldehyde for 15, 20, 30, and 45 min. These results (not shown) indicated that 15 minutes was not sufficient, whereas 45 minutes was.
After washing the 67NR cells, we fixated the cells using paraformaldehyde for 45 minutes and blocked remaining binding sites with 1% BSA. After incubation with plasma of the animals we detected primary antibody binding using an HRP-labelled anti-mouse Ig. In order to validate this cellular ELISA-based protocol, we spiked plasma with the primary H-2K d antibody as described above, followed by titration of the detection antibody with plasma from vaccinated and healthy Balb/C mice (S4a Fig). Using this assay, we demonstrated that seeding of 10 5 cells resulted in a higher signal and less internal variation as compared to seeding of 0.75 *10 5 cells. In addition, we questioned in these experiments whether the fixation protocol could account for background staining. Therefore, we determined the binding of the H-2K d antibody and the secondary antibody Ig-HRP (background) on fixated 67NR cells (15, 20, 30 and 45 min). We observed no increased binding of the H-2K d antibody with longer fixation protocols and even observed less binding of the secondary antibody with longer fixations (data not shown). Subsequently, we determined the optimal dilution for the plasma. Therefore, binding of antibodies of the vaccinated Balb/C mice and control mice was evaluated when plasmas were diluted 20, 40, 80, 160, and 320 times (S4b In order to measure the Ig isotype composition in plasma of tumour-bearing mice we selected antibodies that show minimal crossreactivity with Igs from another subclass. the other subclass specific detection antibodies, which was also confirmed for the other subclasses. In all cases less than 5% of the MFI was observed when the mouse anti-human antibodies were stained with antibodies that are reactive with another subclass (S5c Fig).