Abstract 3916: Patient-derived organoid and cell culture models from the NCI Patient-Derived Models Repository (NCI PDMR) preserve genomic stability and heterogeneity of patient tumor specimens

Background: The National Cancer Institute (NCI) has developed a Patient-Derived Models Repository (PDMR; https://pdmr.cancer.gov) of preclinical models including patient-derived xenografts (PDX), organoids (PDOrg) and patient-derived cell cultures (PDC). Extensive clinical annotation and genomic datasets are available for these preclinical models. However, it is unclear if the molecular profiles of the corresponding patient tumors are stably propagated in these models. We have previously demonstrated that PDX models from the NCI PDMR faithfully represent the patient tumors both in terms of genomic stability and tumor heterogeneity. Here, we conduct an in-depth investigation of genomic representation of patient tumors in the PDOrgs and PDCs.

Methods: PDOrgs (n=64) and PDCs (n=94) were established from tumor fragments (i.e., initiator specimens) obtained either from patient specimens or from PDX specimens of early passage. For some models (n=19), both PDOrgs and PDCs were generated from the same tumor tissue; in fewer cases (n=4), PDCs were established from organoids derived from patient specimens. Whole Exome Sequencing and RNA-Seq were performed on all PDCs and PDOrgs, and data were compared with patient specimens or early passage PDXs.

Results: A majority of the PDOrgs and PDCs have stably inherited the genome of the corresponding patient specimens based on the following observations: (1) >87% of PDOrgs and PDCs maintained similar copy number alteration profiles compared with the initiator specimens of the preclinical model; (2) the variant allele frequency (VAF) of clinically relevant mutations remained consistent between the PDOrgs, PDCs, and the initiator specimens, with none of the PDCs or PDOrgs deviating by >15% VAF; and (3) clinically relevant biomarkers (e.g., MSI, LOH, mutational signatures etc.) are concordant amongst the PDOrgs, PDCs, and the initiator specimens. We observed that the majority of SNVs and indels present in the initiator specimens were also found in the PDOrgs and PDCs, suggesting almost all the tumor heterogeneity was preserved in these preclinical models.

Conclusions: This large and histologically diverse set of PDOrgs and PDCs from the NCI PDMR exhibited genomic stability and faithfully represented the tumor heterogeneity observed in corresponding patient specimens. These preclinical models thus represent a valuable resource for researchers interested in pre-clinical drug or other studies.

Citation Format: Biswajit Das, Yvonne A. Evrard, Li Chen, Rajesh Patidar, Tomas Vilimas, Justine N. McCutcheon, Amanda L. Peach, Nikitha V. Nair, Thomas D. Forbes, Brandie A. Fullmer, Anna J. Lee Fong, Luis E. Romero, Alyssa K. Chapman, Kelsey A. Conley, Robin D. Harrington, Shahanawaz S. Jiwani, Peng Wang, Michelle M. Gottholm-Ahalt, Erin N. Cantu, Gloryvee Rivera, Lindsay M. Dutko, Kelly M. Benauer, Vishnuprabha R. Kannan, Carrie A. Bonomi, Kelly M. Dougherty, Joseph P. Geraghty, Marion V. Gibson, Savanna S. Styers, Abigail J. Walke, Jenna E. Moyer, Anna Wade, Mariah L. Baldwin, Kaitlyn A. Arthur, Kevin J. Plater, Luke Stockwin, Matthew R. Murphy, Michael E. Mullendore, Dianne L. Newton, Melinda G. Hollingshead, Chris A. Karlovich, Paul M. Williams, James H. Doroshow. Patient-derived organoid and cell culture models from the NCI Patient-Derived Models Repository (NCI PDMR) preserve genomic stability and heterogeneity of patient tumor specimens [abstract]. In: Proceedings of the Annual Meeting of the American Association for Cancer Research 2020; 2020 Apr 27-28 and Jun 22-24. Philadelphia (PA): AACR; Cancer Res 2020;80(16 Suppl):Abstract nr 3916.

Dispersal alters soil microbial community response to drought

Microbial communities will experience novel climates in the future. Dispersal is now recognized as a driver of microbial diversity and function, but our understanding of how dispersal influences responses to novel climates is limited. We experimentally tested how the exclusion of aerially dispersed fungi and bacteria altered the compositional and functional response of soil microbial communities to drought. We manipulated dispersal and drought by collecting aerially deposited microbes after precipitation events and subjecting soil mesocosms to either filter-sterilized rain (no dispersal) or unfiltered rain (dispersal) and to either drought (25% ambient) or ambient rainfall for 6 months. We characterized community composition by sequencing 16S and ITS rRNA regions and function using community-level physiological profiles. Treatments without dispersal had lower soil microbial biomass and metabolic diversity but higher bacterial and fungal species richness. Dispersal also altered soil community response to drought; drought had a stronger effect on bacterial (but not fungal) community composition, and induced greater functional loss, when dispersal was present. Surprisingly, neither immigrants nor drought-tolerant taxa had higher abundance in dispersal treatments. We show experimentally that natural aerial dispersal rate alters soil microbial responses to disturbance. Changes in dispersal rates should be considered when predicting microbial responses to climate change.

Gene expression profiling of 49 human tumor xenografts from in vitro culture through multiple in vivo passages--strategies for data mining in support of therapeutic studies

Background

Development of cancer therapeutics partially depends upon selection of appropriate animal models. Therefore, improvements to model selection are beneficial.

Results

Forty-nine human tumor xenografts at in vivo passages 1, 4 and 10 were subjected to cDNA microarray analysis yielding a dataset of 823 Affymetrix HG-U133 Plus 2.0 arrays. To illustrate mining strategies supporting therapeutic studies, transcript expression was determined: 1) relative to other models, 2) with successive in vivo passage, and 3) during the in vitro to in vivo transition. Ranking models according to relative transcript expression in vivo has the potential to improve initial model selection. For example, combining p53 tumor expression data with mutational status could guide selection of tumors for therapeutic studies of agents where p53 status purportedly affects efficacy (e.g., MK-1775). The utility of monitoring changes in gene expression with extended in vivo tumor passages was illustrated by focused studies of drug resistance mediators and receptor tyrosine kinases. Noteworthy observations included a significant decline in HCT-15 colon xenograft ABCB1 transporter expression and increased expression of the kinase KIT in A549 with serial passage. These trends predict sensitivity to agents such as paclitaxel (ABCB1 substrate) and imatinib (c-KIT inhibitor) would be altered with extended passage. Given that gene expression results indicated some models undergo profound changes with in vivo passage, a general metric of stability was generated so models could be ranked accordingly. Lastly, changes occurring during transition from in vitro to in vivo growth may have important consequences for therapeutic studies since targets identified in vitro could be over- or under-represented when tumor cells adapt to in vivo growth. A comprehensive list of mouse transcripts capable of cross-hybridizing with human probe sets on the HG-U133 Plus 2.0 array was generated. Removal of the murine artifacts followed by pairwise analysis of in vitro cells with respective passage 1 xenografts and GO analysis illustrates the complex interplay that each model has with the host microenvironment.

Conclusions

This study provides strategies to aid selection of xenograft models for therapeutic studies. These data highlight the dynamic nature of xenograft models and emphasize the importance of maintaining passage consistency throughout experiments.

Electronic supplementary material

The online version of this article (doi: 10.1186/1471-2164-15-393) contains supplementary material, which is available to authorized users.

Real-time PCR-based assay to quantify the relative amount of human and mouse tissue present in tumor xenografts

Background

Xenograft samples used to test anti-cancer drug efficacies and toxicities in vivo contain an unknown mix of mouse and human cells. Evaluation of drug activity can be confounded by samples containing large amounts of contaminating mouse tissue. We have developed a real-time quantitative polymerase chain reaction (qPCR) assay using TaqMan technology to quantify the amount of mouse tissue that is incorporated into human xenograft samples.

Results

The forward and reverse primers bind to the same DNA sequence in the human and the mouse genome. Using a set of specially designed fluorescent probes provides species specificity. The linearity and sensitivity of the assay is evaluated using serial dilutions of single species and heterogeneous DNA mixtures. We examined many xenograft samples at various in vivo passages, finding a wide variety of human:mouse DNA ratios. This variation may be influenced by tumor type, number of serial passages in vivo, and even which part of the tumor was collected and used in the assay.

Conclusions

This novel assay provides an accurate quantitative assessment of human and mouse content in xenograft tumors. This assay can be performed on aberrantly behaving human xenografts, samples used in bioinformatics studies, and periodically for tumor tissue frequently grown by serial passage in vivo.

STUDY OF THE ACTION OF FOUR AROMATIC CINCHONA DERIVATIVES ON PNEUMOCOCCUS. A COMPARISON WITH OPTOCHIN

It has been shown with one strain of pneumococcus (Type 1, Neufeld), that hydroquinine chloroacetanilide (C 29), hydroquinine p-chloroacetylaminophenol hydrochloride (C 36), hydroquinine m-chloroacetylaminophenol hydrochloride (C 40), and hyroquinine 4-chloroacetylaminopyrocathecol hydrochloride (C 110) have a rapid pneumococcidal activity both in vitro and in the peritoneal cavity of mice, and to a lesser extent in rabbits. In comparison, optochin is slower in action, but its power is not so easily destroyed either in vitro or in vivo. The introduction of the hydroxy group of the benzene nucleus of hydroquinine chloroacetanilide changes the relationship between organotropism and bacteriotropism. In comparing the rapidity of in vitro bactericidal action and intraperitoneal toxicity, C 29 exhibits the most rapid pneumococcidal action and is the most toxic for mice. C 36, the para-hydroxy derivative, is one-fifth as toxic as C 29 and only one-tenth less active bactericidally. C 40 is one-half as toxic and has approximately the same bactericidal power, while C 110 is one-eigth as toxic and one-fifth as pneumococcidal; and optochin is one-sixth as toxic and has one-fifth the bactericidal action. Arranged in the order of their ability to kill pneumococci when injected simultaneously with them into the peritoneal cavity, the drugs are: C 40, C 110, C 36, optochin, and C 29. The chemotherapeutic action of the aromatic compounds is essentially local in character. But by per os therapy there is demonstrated a certain amount of diffusion of this activity, not shown by any other method of administration, C 40 and C 110 having about the same value as optochin. Intravenous injection of the drugs in small doses destroyed to a greater or less extent the natural defenses of the animal, optochin being perhaps less injurious than the aromatic compounds. This same destruction of natural resistance followed intraperitoneal and subcutaneous injections of the chemicals as measured by intravenous injections of the organisms. The maximum tolerant dose in a single injection (intraperitoneal) is not so efficacious as the same dose divided in fifths and injected at hour intervals. Optochin under these conditions is not so active as the aromatic compounds. In general, repeated doses are more curative than single. There is a zone between the therapeutic and toxic doses, both single and repeated, for all these chemicals alike, where the natural resistance of the animal to an infection is reduced. This effect is noted especially with C 29, C 36, and C 40. In the case of optochin the therapeutic dose is nearer the toxic than with C 110, C 36, and C 40. Apparently these chemicals exhibit a variability in in vivo bactericidal activity according to different strains of pneumococci and numerical virulence.

STUDIES ON VIRULENCE : III. INFLUENCE OF HYDROGEN ION CONCENTRATION AND OF INGREDIENTS OF PLAIN BROTH ON THE VIRULENCE OF PNEUMOCOCCI

1. A highly virulent strain of pneumococcus, Type I, Neufeld, when grown in an automatic transferring device at 2 hour intervals on different lots of plain broth adjusted respectively to pH = 6.5, 7, 7.5, or 7.8, lost its virulence at a rate in direct ratio to the H ion concentration of the media-the more acid the media, the more rapid the decrease. 2. Growth of a virulent pneumococcus upon plain broth of an H ion concentration changing gradually either to the acid or the alkaline side of neutrality was accompanied by an initial rise in virulence with a subsequent fall, the change being more rapid in the alkaline than in the acid medium. The organism under these conditions underwent an alteration in its behavior toward agglutinating sera; although still specific, its agglutinability became much greater than that of the original organism. This change was more pronounced in the alkaline than in the acid medium. 3. The virulence of a relatively virulent culture of a pneumococcus grown on plain broth underwent different alterations when the organism was transferred at intervals of 1, 2, 4, or 8 hours. With the 1 hour interval, there occurred seemingly an immediate decrease in virulence, while at intervals of 2, 4, and 8 hours, there occurred first a rise and then a fall, the rise being greatest with the 8 hour interval of transfer and least with the 2 hour. 4. Meat infusion adjusted to various H ion concentrations-pH = 7, 7.3, and 7.7-and made with different amounts of meat furnished conditions which caused a decrease in virulence, the unfavorable action varying in inverse proportion to the concentration of the infusion. Dextrose added to both plain broth and to meat infusion neutralized to a marked degree the unfavorable action of these media. 5. The growth of pneumococci on the different ingredients of plain broth, with plain broth made from the same supply as a control, revealed the fact that peptone in 2 per cent solution maintained and even increased the virulence of the strain of pneumococcus studied, while meat infusion caused the usual decrease. The rate of decrease with plain broth was slower than with meat infusion, showing that the 2 per cent peptone neutralized to some extent the unfavorable action of the infusion.

STUDIES ON VIRULENCE : II. THE INCREASE IN VIRULENCE IN VITRO OF A STRAIN OF PNEUMOCOCCUS

1. An automatic apparatus is described by which a new supply of food can be furnished actively growing organisms at any desired interval of time-an automatic transferring device. 2. A single strain of pneumococcus, Type I, Neufeld, which had become avirulent for mice, acquired virulence of maximal degree when grown in the described apparatus with skimmed milk sterilized in an Arnold sterilizer as the medium. Transfers were made at intervals of 2, 4, and 8 hours. It is difficult to determine the best interval of transfer, but the 8 hour interval apparently is most suitable, with the 4 and 2 hour following in preference in the order named. 3. Pasteurized skimmed milk from a single dairy, but obtained on different days, when used as medium at a 2 hour interval, varied in effect on the virulence of pneumococci, the results showing that the virulence might be either increased, decreased, or maintained. 4. Skimmed milk heated for varying lengths of time, 30 minutes at 17 pounds pressure, or 60 or 90 minutes at the same pressure, lost its suitability for maintaining virulence of a pneumococcus when transferred every 2 hours, the effect being in direct proportion to the length of time the milk was heated. 5. The H ion concentration of milk had slight effect on virulence. A virulent strain of pneumococcus was grown in milk adjusted to pH = 5, 6, 7, 8, and 9, transfers being made every 2 hours. Virulence was maintained to a like degree on milk titrated to pH = 5, 6, and 7, but when the organism was grown in milk of pH = 8 and 9, virulence decreased, more rapidly at pH = 9 than at pH = 8. 6. A pure line practically avirulent strain of pneumococcus picked by the Barber method, when grown in milk at a 4 hour interval of transfer, increased in virulence 10 million fold; that is to say, until one diplococcus would kill a mouse. Prior to this study, no record has been found in which the virulence of any microorganism has been increased to such a degree by an in vitro method.

Parathyroid hormone-related protein as a growth regulator of prostate carcinoma

Parathyroid hormone-related protein (PTHrP) is produced by prostate carcinoma cells and tumors, but little is known of its role in prostate carcinogenesis. The goal of this study was to evaluate PTHrP expression in the regulation of prostate carcinoma growth using human and animal models. PTHrP expression was assessed in prostate cancer cell lines in vitro. Seven of nine cell lines produced PTHrP, and increased expression was seen during cell proliferation. The MatLyLu rat prostate carcinoma model was used to determine the effects of PTHrP overexpression on prostate tumor growth. PTHrP overexpression did not alter proliferation of the cells in vitro. However, when PTHrP-overexpressing cells were injected into rat hind limbs, primary tumor growth and tumor size were significantly enhanced as compared with control cells. To evaluate PTHrP in human prostate carcinoma patients, immunohistochemistry was performed on metastatic bone lesions. Immunolocalization of PTHrP protein was found in the cytoplasm and nucleus of cancer cells in the bone microenvironment. Because nuclear localization of PTHrP has been associated with an inhibition of apoptosis, the ability of full-length PTHrP to protect prostate cancer cells from apoptotic stimuli was examined. Cells transfected with full-length PTHrP showed significantly increased cell survival after exposure to apoptotic agents as compared with cells producing no PTHrP (plasmid control) or cells transfected with PTHrP lacking its nuclear localization signal. To determine the mechanism of action of PTHrP in prostate cancer cells, the parathyroid hormone/PTHrP receptor status of the cells was determined. These cell lines did not demonstrate parathyroid hormone/PTHrP receptor-mediated binding of iodinated PTHrP or steady-state receptor message by Northern blot analysis, but they did have a detectable receptor message by reverse transcription-PCR analysis. In summary, PTHrP is expressed in many prostate cancer cell lines in vitro and in metastatic bone lesions in vivo. PTHrP expression positively influences primary tumor size in vivo and protects cells from apoptotic stimuli. These data suggest that PTHrP plays an important role in the promotion of prostate tumor establishment and/or progression.

Skeletal metastasis of prostate adenocarcinoma in rats: morphometric analysis and role of parathyroid hormone-related protein

BACKGROUND

Prostate cancer frequently metastasizes to bone, where it induces osteoblastic lesions. Parathyroid hormone-related protein (PTHrP), a product of normal and neoplastic prostate cells, may promote growth and bone metastasis of certain types of cancer. In this study, we investigated the: 1) pathogenesis and morphology of bone metastases in the MATLyLu rat prostate adenocarcinoma model, and 2) effect of PTHrP overexpression on tumor growth and incidence of bone metastasis.

METHODS

MATLyLu cells were stably transfected with a PTHrP expression vector or control plasmid. PTHrP expression was determined in vitro by immunoradiometric assay and Northern blot analysis. MATLyLu cells were injected into the left ventricle of Copenhagen rats to induce bone metastases. Histology and radiography were used to quantify the size and number of bone metastases. Serum alkaline phosphatase isoenzyme concentrations and histomorphometric analysis were used to evaluate bone formation and resorption.

RESULTS

All rats developed osteolytic metastases in long bones and vertebrae. There was no evidence of increased intramedullary bone formation. PTHrP overexpression by MATLyLu cells was not associated with any difference in the incidence of bone metastasis, size of metastatic foci or tumor-cell proliferation.

CONCLUSIONS

The MATLyLu intracardiac injection model of prostate carcinoma is an aggressive tumor model with a high incidence of osteolytic skeletal metastases, and is not altered by increased PTHrP production by neoplastic prostate epithelial cells.

The plasminogen activator inhibitor-2 gene is not required for normal murine development or survival

Plasminogen activator inhibitor-2 (PAI-2), a member of the serpin gene family, is thought to serve as a primary regulator of plasminogen activation in the extravascular compartment. High levels of PAI-2 are found in keratinocytes, monocytes, and the human trophoblast, the latter suggesting a role in placental maintenance or embryo development. The primarily intracellular distribution of PAI-2 also may indicate a unique regulatory role in a protease-dependent cellular process such as apoptosis. To examine the potential functions of PAI-2 in vivo, we generated PAI-2-deficient mice by gene targeting in embryonic stem cells. Homozygous PAI-2-deficient mice exhibited normal development, survival, and fertility and were also indistinguishable from normal controls in response to a bacterial infectious challenge or endotoxin infusion. No differences in monocyte recruitment into the peritoneum were observed after thioglycollate injection. Epidermal wound healing was equivalent among PAI-2 -/- null and control mice. Finally, crossing PAI-2 -/- with PAI-1 -/- mice to generate animals deficient in both plasminogen activator inhibitors failed to uncover an overlap in function between these two related proteins.

Expression and processing of human ornithine-delta-aminotransferase in Saccharomyces cerevisiae

Ornithine-delta-aminotransferase catalyzes the conversion of ornithine to glutamate-gamma-semialdehyde. In humans, deficiency of this mitochondrial matrix enzyme results in the progressive blinding disorder, gyrate atrophy of the choroid and retina. To explore yeast as an expression system, we introduced a cDNA encoding human ornithine-delta-aminotransferase into an ornithine aminotransferase-deficient strain of Saccharomyces cerevisiae. The human enzyme was expressed at high levels, with activity 20-fold greater than that of wild-type yeast and 10-fold higher than in human fibroblasts. Although the normal location of ornithine-delta-aminotransferase in S. cerevisiae is cytosolic, human ornithine-delta-aminotransferase expressed in S. cerevisiae was localized to the mitochondrial matrix with correct proteolytic processing of its mitochondrial leader sequence. Despite this anomalous location in yeast, human ornithine-delta-aminotransferase complemented the phenotype of the mutant strain, restoring its ability to utilize ornithine as a sole nitrogen source. We also expressed a vitamin B6-responsive missense allele of ornithine-delta-aminotransferase (V332M) and showed that the biochemical phenotype of this allele is easily demonstrated confirming the usefulness of this system for examining mutations causing gyrate atrophy.

Cloning human pyrroline-5-carboxylate reductase cDNA by complementation in Saccharomyces cerevisiae

Pyrroline-5-carboxylate reductase (EC 1.5.1.2) catalyzes the NAD(P)H-dependent conversion of pyrroline-5-carboxylate to proline. We cloned a human pyrroline-5-carboxylate reductase cDNA by complementation of proline auxotrophy in a Saccharomyces cerevisiae mutant strain, DT1100. Using a HepG2 cDNA library in a yeast expression vector, we screened 10(5) transformants, two of which gained proline prototrophy. The plasmids in both contained similar 1.8-kilobase inserts, which when reintroduced into strain DT1100, conferred proline prototrophy. The pyrroline-5-carboxylate reductase activity in these prototrophs was 1-3% that of wild type yeast, in contrast to the activity in strain DT1100 which was undetectable. The 1810-base pair pyrroline-5-carboxylate reductase cDNA hybridizes to a 1.85-kilobase mRNA in samples from human cell lines and predicts a 319-amino acid, 33.4-kDa protein. The derived amino acid sequence is 32% identical with that of S. cerevisiae. By genomic DNA hybridization analysis, the human reductase appears to be encoded by a single copy gene which maps to chromosome 17.

An RNA polymerase I promoter located in the CHO and mouse ribosomal DNA spacers: functional analysis and factor and sequence requirements

We report results of experiments in which we demonstrated the existence of a polymerase I promoter within the ribosomal DNA spacer upstream from the rRNA initiation site in Chinese hamsters and mice. Transcription of the CHO spacer promoter was achieved by the same protein factors, C and D, that catalyzed transcription of the gene promoter, and these factors bound stably to the CHO spacer promoter in a preinitiation complex, just as they did to the gene promoter. In contrast to the CHO spacer promoter, which was transcribed in vitro nearly as efficiently as the gene promoter, the mouse spacer promoter was far less active; this low activity was attributable to the fact that the mouse spacer promoter bound factor D inefficiently. It is striking that the active CHO spacer promoter violated the otherwise universal rule that metazoan RNA polymerase I promoters all have a G residue at position -16. Sequence comparisons also revealed a great similarity between the CHO and mouse spacer promoter regions, yet there was much less similarity between the flanking sequences. There was also only limited homology between the spacer and gene promoter regions, but despite this the two kinds of initiation regions were organized similarly, both consisting of an essential core promoter domain and a stimulatory domain that extended upstream to approximately residue -135. Evolutionary considerations argue strongly that the presence of ribosomal DNA spacer promoters offers a significant selective advantage.

THE ORGANOTROPIC, BACTERIOTROPIC, AND LEUCOCYTOTROPIC ACTIONS OF CERTAIN ORGANIC CHEMICALS

We are dealing, as the results show, with groups of chemicals, all of which, whether bacteriotropic or not, greatly inhibit the engulfing of Staphylococcus aureus by leucocytes. Not a sufficiently large number of experiments was performed in attempt to cure experimental staphylococcus infections to warrant any condusion in regard to possible therapeutic activity against this organism. How-ever, as will appear in another paper, the only group out of the seven which definitely possessed an in vivo bactericidal action against pneumococcus is that of the cinchona derivatives. Certain members of the other chemical groups studied, although bactericidal in a very high dilution, -chemicals in which the concentration of a non-lethal dose was many times greater than that required to kill multiple minimal lethal doses of organisms in vitro, -had no certain effect when bacteria and drug were injected simultaneously into the peritoneal cavity of a mouse. In fact, the treated mouse often died before the controls. If we may assume,-leaving out of consideration the practical significance of in vivo chemical destruction and excretion following the injection of the drug into the animal,-that the failure of these chemicals to exhibit a benign influence on a systemic infection in cases in which the drug can be used in a bactericidal dilution, is due to their antiphagocytic property, only one step has been taken in analysis of the factors vital for the defense of the animal against a specific microorganism. Why do these chemicals inhibit leucocytic activity? Is it because of their influence upon complement, opsonin, or the leucocyte itself, or some special one function that determines the ability to ingest bacteria? Only further work can definitely settle this question and also determine whether or not such an analysis would be of practical importance in a rational development of chemotherapy. The ideal chemotherapeutic agent may be one that has an in vivo bactericidal potency and a negligible or stimulatory phagocytic action in doses non-lethal for the experimental animal. However difficult such a drug may be to find, it seems unlikely that the ultimate success in chemotherapy will be so simple. Again, it is conceivable that a secondary action of a drug, although leucocytotropic and not bacteriotropic, may bring about conditions in the animal body that will enable it to throw off the invading organism. Or finally, a drug compatible with the forces necessary to the host's defense and possessing in vivo bactericidal action to a greater or less degree may be the chemical sought for, the goal toward which we should strive, to achieve a rational chemotherapy for infectious diseases.