Research Article
`
`Received: 23 December 2016
`
`Revised: 13 April 2017
`
`Accepted article published: 9 May 2017
`
`Published online in Wiley Online Library: 12 August 2017
`
`(wileyonlinelibrary.com) DOI 10.1002/ps.4610
`
`Cyclohexylamine inexplicably induces
`antennae loss in Formosan subterranean
`termites (Coptotermes formosanus Shiraki):
`cyclohexylamine hydrogen phosphate salts are
`novel termiticides
`Brantley Grimball,a,† Lucas Veillon,a,b,c,† Tara Calhoun,a Frank R Fronczek,d
`Emily Arceneauxa and Roger A Lainea,d*
`
`Abstract
`
`Shirakii),
`(Coptotermes
`termites
`subterranean
`Formosan
`experiments with
`In
`BACKGROUND:
`formosanus
`myo-inositol-2-monophosphate as the dicyclohexylammonium salt was tested among other sugar derivatives, and was
`found to be toxic to C. formosanus when added to a moistened filter paper food source in plastic Petri dishes.
`
`RESULTS: Curiously, over a nine-day period, the moniliform (beaded) antenna of C. formosanus deteriorated in a stepwise fash-
`ion with the most distal pseudosegment (bead) turning brown and falling off, followed by the penultimate pseudosegment,
`sequentially, until 7–9 days when only a stub of the antenna remained. Termites became increasingly moribund with the loss
`of antennae, and quit normal behavior including consuming cellulose food, and died. sn-Glycerol-3-phosphate as the dicyclo-
`hexylammonium salt also gave the same results. Dicyclohexylammonium hydrogen phosphate and monocyclohexylammonium
`dihydrogen phosphate were synthesized, to find a low-cost form for application to baits, both of which also showed similar toxi-
`city. In a trial with Fibonacci series dilutions of neat cyclohexylamine, the antenna-affecting activity became apparent in the LD30
`(14 days) to LD70 range of concentrations. At the higher concentrations, darkening of the most distal parts of leg extremities was
`noticed.
`
`CONCLUSION: Cyclohexylamine appears to be a novel termiticide with a previously unreported mechanism of toxicity. Its
`hydrogen phosphate salts retain the toxic effect and are inexpensive and easily synthesized.
`© 2017 Society of Chemical Industry
`
`Keywords: cyclohexylamine; termiticide; Formosan termite; antennae
`
`2039
`
`1 INTRODUCTION
`Coptotermes formosanus Shiraki, the Formosan subterranean ter-
`mite, is a cellulophage, endemic in Asia, and exotic in the southern
`USA, California and Hawaii. Formosan termites infest and damage
`wooden structures and more than 50 species of living plants,1– 4
`and can also physically damage non-cellulosic materials such as
`insulation on buried electrical and telephone wires.
`In a series of experiments to examine the effect of rare sug-
`ars (sugars not found free or in abundance in nature) on
`termite and hindgut protozoan survival, one example was
`myo-inositol-2-monophosphate (I2P), the product of phytase
`on phytic acid.5 Although about seven rare sugars showed signif-
`icant toxicity when consumed by adding to paper food sources,
`including 2-deoxy-galactose,6,7 I2P had a very unusual toxicity
`profile, causing deterioration of sequential pseudosegments of
`antennae over a 9-day period during a 14-day LD50 test screening.
`The deteriorating antennae phenomenon was noticed by under-
`graduate co-author TC while dissecting I2P-treated C. formosanus
`hindgut under a microscope for symbiotic protozoa counts
`
`on a hemocytometer. This observation was further examined
`with I2P concentration–response experiments, and confirmed
`in repeated experiments. While I2P itself was thought to be the
`
`∗ Correspondence to: RA Laine, Departments of Biological Sciences and Chem-
`istry, Louisiana State University and A&M College, Baton Rouge, LA 70803, USA.
`E-mail: rlaine@lsu.edu
`
`† These authors contributed equally to this study.
`
`a Department of Biological Sciences, Louisiana State University and A&M College,
`Baton Rouge, LA, USA
`
`b Division of Glycopathology, Institute of Molecular Biomembrane and Glycobi-
`ology, Tohoku Pharmaceutical University, Sendai, Japan
`
`c Department of Chemistry and Biochemistry, Texas Tech University, Lubbock, TX,
`USA
`
`d Department of Chemistry, Louisiana State University and A&M College, Baton
`Rouge, LA, USA
`
`Pest Manag Sci 2017; 73: 2039–2047
`
`www.soci.org
`
`© 2017 Society of Chemical Industry
`
`Merck Exhibit 2203, Page 1
`Mylan Pharmaceuticals Inc. v. Merck Sharp & Dohme Corp.
`IPR2020-00040
`
`

`

`www.soci.org
`
`B Grimball et al.
`
`Figure 1. Means (±SE) of C. formosanus mortality following treatment of
`food source filter paper with I2P (dicyclohexylammonium salt) in no-choice
`assays with 20 workers (n = 3). Collection group B was used and the data
`were recorded daily for 14 days. The F value for percent mortality is 54.23.
`P and df values are <0.0001 and 74, 150.
`
`Figure 2. Chemical structure of I2P.
`
`2.3 Crystal structure determination
`Crystal structures were determined by co-author FRF using X-ray
`diffraction data measured at low temperature (90 K) with a Bruker
`Kappa Apex-II DUO diffractometer. Hydrogen atoms were located,
`and positions for those on N and O were refined, demonstrating
`that, in both cases, the cyclohexylamine exists as the ammonium
`cation. For the monocyclohexylammonium salt, the phosphate
`−; for the 2:1
`exists as dihydrogen phosphate monoanion, H2PO4
`2−.
`salt, the phosphate exists as hydrogen phosphate dianion, HPO4
`
`2.4 Supporting information available
`Crystallographic Information Files for the crystal structures have
`been deposited at the Cambridge Crystallographic Data Centre as
`CCDC 1452252 and 1452253 and can be obtained free of charge
`(www.ccdc.cam.ac.uk/data_request/cif). Data collection: Bruker
`APEX2; cell refinement: Bruker SAINT; data reduction: Bruker
`SAINT; programs used to solve structure: SHELX979; programs
`used to refine structure: SHELX979; molecular graphics: ORTEP-3
`for Windows10; software used to prepare material for publication:
`SHELXL97.9
`
`2.5 Concentration–response feeding assays
`I2P (dicyclohexylammonium salt), glycerol-3-phosphate (dicy-
`clohexylammonium salt), dicyclohexylammoiumn hydrogen
`
`culprit molecule, its unusual dicyclohexylammonium salt form was
`noticed and a completely unrelated sugar was tested as the same
`salt, sn-glycerol-3-phosphate dicyclohexylammonium. It, too, had
`the same unusual antennae deterioration effect on the termites,
`suggesting that dicyclohexylammonium salt itself conferred the
`toxic effect. Next, crystalline dicyclohexylammonium hydrogen
`phosphate was prepared and assayed using X-ray diffraction for
`confirmation of the structure. Dicyclohexylammonium hydrogen
`phosphate had a toxicity profile similar to that of the sugar deriva-
`tives. Upon performing a Fibonacci series water dilution trial of
`neat cyclohexylamine, which was immediately toxic to termites in
`higher concentrations, the 14-day LD50 was determined. At levels
`between LD70 (2.5 ng μL−1) and LD30 (1.5 ng μL−1), where termites
`survived for a few days, the antennae deterioration effect was
`observed. In addition, the most distal extremities of the legs also
`showed darkening, suggesting as one possibility that some inhi-
`bition of the insect hemolymph circulatory system was involved.
`Therefore cyclohexylamine itself was responsible for all of the
`morphological changes and termite toxicity of these compounds.
`
`2 MATERIALS AND METHODS
`2.1 Insect collection and maintenance
`Worker C. formosanus were collected from Brechtel Park, New
`Orleans, Louisiana by methods of Smith et al.8 Termites were
`collected from New Orleans on 13 June 2006 (collection group
`A); 9 August 2006 (collection group B); 1 March 2008 (collection
`group C); 9 May 2008 (collection group D); 20 May 2008 (col-
`lection group E); 20 March 2009 (collection group F); 29 May
`2009 (collection group G); 11 May 2010 (collection group H); 18
`August 2012 (collection group I); May 2014 (collection group J);
`and August 2014 (collection group K). The colonies were main-
`tained with infested and new wood in garbage cans set in tubs
`with a few inches of water. Some termite collection groups had
`been in culture for a few months before experiments, and some
`were used within a month of collection. The termite collection
`procedure was tapping infested wood sticks into clean plastic
`containers where termites were collected on moistened paper
`towels.
`
`2.2 Chemicals
`I2P (dicyclohexylammonium salt), glycerol-3-
`Neutral red dye,
`phosphate (dicyclohexylammonium salt) and cyclohexylamine
`were obtained from Sigma-Aldrich (St Louis, MO). D-Galactose
`and D-glucose were obtained from Matheson Coleman and Bell
`(Cincinnati, OH) and Fisher Scientific (Fair Lawn, NJ), respectively.
`Dicyclohexylammonium hydrogen phosphate was synthesized
`by titrating 85% phosphoric acid into a 1 M solution of cyclo-
`hexylamine in a 2:1 molar ratio of neat cyclohexylamine and 85%
`phosphoric acid solution. After being allowed to react for 24 h, the
`resulting solution was stored in a glass beaker for approximately
`one week at room temperature until large crystals formed. The
`pinkish crystalline product was gently scraped from the beaker
`and collected for use. Monocyclohexylammonium dihydrogen
`phosphate was synthesized by carefully titrating 85% phospho-
`ric acid into neat cyclohexylamine in an ice bath in a 1:1 molar
`ratio. The crystals were clusters of needles and appeared white.
`Analysis using X-ray crystallography elucidated and confirmed the
`identity of the product crystals for both mono- and dicyclohexy-
`lammonium hydrogen phosphate that were subsequently used
`for testing.
`
`2040
`
`wileyonlinelibrary.com/journal/ps
`
`© 2017 Society of Chemical Industry
`
`Pest Manag Sci 2017; 73: 2039–2047
`
`Merck Exhibit 2203, Page 2
`Mylan Pharmaceuticals Inc. v. Merck Sharp & Dohme Corp.
`IPR2020-00040
`
`

`

`Cyclohexylamine induces antennae loss in Formosan termites
`
`www.soci.org
`
`Figure 3. Photographs showing a control termite fed untreated filter paper and termites after feeding on I2P (dicyclohexylammonium salt)-treated filter
`paper for three days and nine days (mortality is 60% at day 9, this shows a surviving termite). Doses were 1281.7 and 640.8 μg mm−3, respectively.
`
`2041
`
`hydrogen phosphate and monocyclohexylammonium dihydro-
`gen phosphate were tested for toxicity. Dicyclohexylammmonium
`hydrogen phosphate was tested in concentrations ranging from
`0.348 to 348 μg mm−3, and monocyclohexylammmonium dihy-
`drogen phosphate was tested in concentrations ranging from
`2.28 to 228 μg mm−3. Replicates of twenty worker termites were
`incubated in the dark in a Parafilm-sealed Petri dish at room
`temperature for two weeks. Termite mortality was recorded daily
`in triplicate experiments. An analysis of variance (ANOVA; SAS
`ANOVA procedure) followed by Tukey’s Studentized range test
`was used to evaluate statistical differences among groups.11 All
`mortality data were judged at 𝛼 = 0.05.
`Termites from collection group B were used in the I2P con-
`centration mortality assay. Termites from collection groups C
`and D were used in both the dicyclohexylammonium-I2P and
`dicylclohexylammonium-glycerol-3-phosphate
`concentration
`mortality assay. Termites from collection group I were used in
`cyclohexylamine concentration mortality assays. Termites from
`collection group J were used in the dicyclohexylammonium
`hydrogen phosphate and neat cyclohexylamine volatility concen-
`tration mortality assays. Termites from collection group K were
`used in the monocyclohexylammonium dihydrogen phosphate
`concentration mortality assay.
`In concentration mortality experiments where paper consump-
`tion data are shown, filter papers were weighed in ambient humid-
`ity prior to sugar application. After 14 days, termites were removed
`from the dishes and the filter paper from each replicate was
`cleaned, washed of residual carbohydrate and dried at 100 ∘C for
`24 h. After drying and humidity equilibration for 24 h, filter papers
`were reweighed for determination of paper consumption. ANOVA
`(SAS ANOVA procedure) followed by Tukey’s Studentized range
`test was used to evaluate statistical differences among groups.11
`All consumption data were judged at 𝛼 = 0.05.
`
`2.6 Termite antenna ablative shortening induced by I2P as
`dicyclohexylammonium salt
`Fifty worker termites, from collection group B, were placed in a
`60 mm × 15 mm polystyrene Petri dish with 42.5 mm filter paper
`treated with 10 mg of I2P (640.8 μg mm−3). Daily, for 14 days, three
`randomly selected termites were removed, euthanized with 5 μL of
`ethanol and their antennae were photographed, after which GSA
`Image Analyzer was used to quantify termite antennae length.12
`Antenna length data were subjected to ANOVA followed by Tukey’s
`Studentized range test.11 All data were judged at 𝛼 = 0.05. All
`experiments were performed in triplicate.
`
`Figure 4. Means (±SE) of C. formosanus antenna length. The antennae of
`workers from collection group B were measured daily for 14 days while
`termites were allowed to feed on filter paper treated with 640.8 μg mm−3
`of I2P. The F value for antenna length is 16.90. P and df values are <0.0001
`and 29, 510.
`
`phosphate, monocyclohexylammonium dihydrogen phosphate
`and neat cyclohexylamine were screened for termiticidal activity
`as follows. Test compounds were applied to 42.5 mm filter papers,
`in 60 mm × 15 mm plastic Petri dishes, in a solution of 1 mg of test
`compound per 10 μL of distilled water (dH2O). Doses of 2.5 mg
`(160.2 μg mm−3 of filter paper), 5 mg (320.4 μg mm−3), 10 mg
`(640.8 μg mm−3) and 20 mg (1281.6 μg mm−3) of I2P were exam-
`ined. Following filter paper treatment, dH2O was added to bring
`the total volume applied to the filter papers to 300 μL, using 300 μL
`of dH2O as control. Cyclohexylamine was applied to the paper
`diluted in 300 μL of dH2O, in concentrations ranging from 0.626
`to 10.2 μg mm−3. To prevent desiccation, approximately 50 μL of
`dH2O was applied to filter papers throughout the trials every third
`day, or upon observing that a filter paper disc was relatively dry.
`To test whether termite mortality was induced by consumption
`of the cyclohexylamine-treated paper, or whether the presence
`of cyclohexylamine vapor was responsible for the toxicity, we
`compared addition of cyclohexylamine to both paper and fiber-
`glass filters – the latter of which termites will not consume.
`In addition to pure cyclohexylamine, dicyclohexylammonium
`
`Pest Manag Sci 2017; 73: 2039–2047
`
`© 2017 Society of Chemical Industry
`
`wileyonlinelibrary.com/journal/ps
`
`Merck Exhibit 2203, Page 3
`Mylan Pharmaceuticals Inc. v. Merck Sharp & Dohme Corp.
`IPR2020-00040
`
`

`

`www.soci.org
`
`B Grimball et al.
`
`Figure 5. Means (±SE) of C. formosanus hindgut protozoan counts. (A) Total protozoan counts; (B) Spirotrichonympha, (C) Pseudotrichonympha; (D)
`Holomastigotoides. The hindgut protozoa of workers from collection group A were enumerated daily for 14 days while termites were allowed to feed
`on filter paper treated with 640.8 μg mm−3 of I2P (dicyclohexylammonium salt). The F values for total protozoa, Pseudotrichonympha, Holomastigotoides
`and Spirotrichonympha are 22.45 (P < 0.0001), 19.62 (P < 0.0001), 26.21 (P < 0.0001) and 6.83 (P < 0.0001), respectively. The df value is 19, 20 for all.
`
`2.7 Protozoan counting
`The effect of I2P dicyclohexylammonium salt on C. formosanus
`hindgut protozoan populations was examined as follows. An
`amount of 10 mg of I2P was dissolved in 200 μL of dH2O and
`in a 60 mm × 15 mm plastic Petri dish
`applied to filter paper,
`(640.8 μg mm−3). Control filter paper received 200 μL of dH2O. With
`75 worker termites per dish, collection group A was used. Pseu-
`dotrichonympha grassii Koidzumi, Holomastigotoides hartmanni
`Koidzumi and Spirotrichonympha leidyi Koidzumi were counted
`daily for two weeks as described by Mannesmann13 and modi-
`fied by Maistrello et al.14 Hindguts were removed from the pos-
`terior ends of three workers and gently macerated in 40 μL of
`saline containing neutral red dye (0.5 mL of 1% aqueous neu-
`tral red solution dissolved in 10 mL of saline). The number of
`each protozoan species was determined using a hemocytome-
`ter (Bright-line Improved Neubauer, Hausser Scientific, Horsham,
`PA) under a light microscope. The population of each protozoan
`species per hindgut (XF) was calculated as XF = (G × n)/(V × 3),
`where G is the volume (μL) of solution in which hindguts were dis-
`sected, n the mean of two counts within the hemocytometer and
`V the volume of area counted. Mean (±SE) XF values calculated
`from two replicates were used for graphical comparison of data.
`
`Protozoan population data were subjected to ANOVA, followed by
`Tukey’s Studentized range test.11 All data were judged at 𝛼 = 0.05.
`A square root transformation was applied for data analysis; how-
`ever, untransformed means are reported.
`
`3 RESULTS
`3.1 Concentration–response feeding assays
`I2P cyclohexylammonium salt (structure of I2P is shown in Fig. 2)
`(1281.7, 640.8, 320.4 and 160.2 μg mm−3). Concentration-dependent
`toxicity. In this assay the three lowest concentrations applied,
`640.8, 320.4 and 160.2 μg mm−3, did not result in mortality signifi-
`cantly different from the control group; data collected for the two
`lowest concentrations, 320.4 and 160.2 μg mm−3, are not shown
`(Fig. 1). Mortality observed in the highest dosage group, 1281.7 μg
`mm−3, became significant on day eight and thereafter (Fig. 1). A
`mortality of 43% was observed on day eight, increasing linearly to
`98% on day 14 (Fig. 1).
`
`3.2 Quantifying termite antenna length
`I2P (dicyclohexylammonium salt) (640.8 μg mm−3). Distal pseudoseg-
`ment shortening. A 640.8 μg mm−3 I2P (dicyclohexylammonium
`
`wileyonlinelibrary.com/journal/ps
`
`© 2017 Society of Chemical Industry
`
`Pest Manag Sci 2017; 73: 2039–2047
`
`2042
`
`Merck Exhibit 2203, Page 4
`Mylan Pharmaceuticals Inc. v. Merck Sharp & Dohme Corp.
`IPR2020-00040
`
`

`

`Cyclohexylamine induces antennae loss in Formosan termites
`
`www.soci.org
`
`Figure 6. Means (±SE) of C. formosanus mortality following treatment of
`food source filter paper with glycerol-3-phosphate dicyclohexylamine salt
`in no-choice assays with 20 workers (n = 3). Collection group D was used
`and the data were recorded daily for 14 days. The F value for percent
`mortality is 9.163. P and df values are <0.0001 and 2, 42.
`
`Figure 9. Means (±SE) of C. formosanus mortality following treatment
`of food source filter paper with neat cyclohexylamine in two separate
`no-choice assays with 20 workers (n = 3). Collection group I was used and
`the data were recorded daily for 14 days. The F value for percent mortality
`(1.0 and 1.7) is 11.13. P and df values are <0.0001 and 3, 56. The F value for
`percent mortality (2.4) is 11.13. P and df values are <0.0001 and 4, 70.
`
`Figure 7. Chemical structure of cyclohexylamine.
`
`2043
`
`Figure 10. Means (±SE) of C. formosanus filter paper consumption over
`14 days. Food source filter paper was treated with neat cyclohexylamine
`before incubation. Collection group I was used and consumption data were
`recorded at the end of the two week assay. The F and P values are 15.30
`and 0.0003. Data were standardized to live termites through the 14-day
`study.
`
`3.4 Concentration–response feeding assays
`Glycerol-3-phosphate dicyclohexylammonium salt (19.58, 39.15 μg
`mm−3). Concentration-dependent toxicity. Both concentrations
`tested resulted in mortality significantly different from the control
`(Fig. 6). In the higher dosage group, mortality became signifi-
`cant on day 5 and increased in a linear fashion, reaching 98%
`mortality on day 12 (Fig. 6). In the lower dosage group, mortality
`became significant on day 12 and increased to 71% on day 14
`(Fig. 6).
`
`3.5 Concentration–response feeding assays
`10.2 μg
`mm−3).
`Neat
`cyclohexylamine
`(3.39,
`5.08,
`Concentration-dependent toxicity. The structure of cyclohexy-
`lamine is shown in Fig. 7. All dosage groups displayed 100%
`mortality within one to three days. 100% mortality was observed
`on day one of the highest dosage, 10.16 μg mm−3, and on day
`three of the lowest dosage, 3.39 μg mm−3 (Fig. 8).
`
`Figure 8. Means (±SE) of C. formosanus mortality following treatment of
`food source filter paper with neat cyclohexylamine in no-choice assays with
`20 workers (n = 3). Collection group I was used and the data were recorded
`daily for 14 days. The F value for percent mortality is 34.66. P and df values
`are <0.0001 and 3, 56.
`
`salt) treatment resulted in significantly degraded termite antennae
`from day five and six and eight through 14 (Figs 3 and 4).
`
`3.3 Protozoa quantification
`I2P bioassay (640.8 μg mm−3). Reduced populations. A 640.8 μg
`mm−3 treatment of I2P (dicyclohexylammonium salt) resulted in
`reduced total protozoan populations on day four and thereafter
`during the 14-day assay (Fig. 5). H. hartmanni populations were sig-
`nificantly reduced on days 1, 2, 5, 8, 9, 10, 11 and 14 (Fig. 5). P. grassii
`populations were also significantly lowered on days 1, 5, 8, 9, 10
`and 11 (Fig. 5). S. leidyi populations were affected the least, only
`being significantly lower than controls on days 9 and 11 (Fig. 5).
`
`Pest Manag Sci 2017; 73: 2039–2047
`
`© 2017 Society of Chemical Industry
`
`wileyonlinelibrary.com/journal/ps
`
`Merck Exhibit 2203, Page 5
`Mylan Pharmaceuticals Inc. v. Merck Sharp & Dohme Corp.
`IPR2020-00040
`
`

`

`www.soci.org
`
`B Grimball et al.
`
`Figure 11. Means (±SE) of C. formosanus mortality following treatment of food source cellulose filter paper and fiberglass filter paper with neat
`cyclohexylamine in no-choice assays with 20 workers (n = 3). Collection group J was used and the data were recorded daily for 14 days. The F value
`for percent mortality is 25.76. P and df values are <0.0001 and 3, 56.
`
`3.7 Cyclohexylamine volatile effects
`Neat cyclohexylamine (600 μg). Consumption-dependent toxicity. In
`this assay, neat cyclohexylamine treatment of cellulose resulted
`in significantly different mortality from the corresponding control
`group on day eight and thereafter (Fig. 11). Twenty-five percent
`mortality was observed on day eight, increasing to 45% on day
`14 (Fig. 11). Cyclohexylamine treatment of fiberglass did not result
`in mortality significantly different from that of the corresponding
`control group (Fig. 11). Since cyclohexylamine is volatile, appar-
`ently consumption is necessary for the toxicity, not just exposure
`to vapors which would have been similar for cellulose and fiber-
`glass. Mortality for the fiberglass control and cellulose control was
`not observed to be significantly different (Fig. 11).
`
`3.8 Concentration–response feeding assays
`Dicyclohexylammonium hydrogen phosphate (17.22, 34.44, 68.88 μg
`mm−3). Concentration-dependent toxicity. All three concentration
`groups were observed to be significantly different from the control
`group (Fig. 12). The highest final mortality value was observed in
`the highest dosage group, 68.88 μg mm−3 (Fig. 12). In this group,
`mortality became significantly different on day five and increased
`in linear fashion to 45% mortality on day 14 (Fig. 12).
`Dicyclohexylammonium hydrogen phosphate (1.22, 2.45, 4.90,
`9.80 μg mm−3). Concentration-dependent toxicity. The two highest
`dosage groups, 4.90 and 9.80 μg mm−3, displayed mortality sig-
`nificantly different from the control (Fig. 13). In the 4.90 μg mm−3
`dosage group, mortality became significantly different on day
`three and increased to 38% on day 14. In the 9.8 μg mm−3 dosage
`group, mortality became significantly different on day four and
`increased to 36% on day 14.
`Dicyclohexylammonium hydrogen phosphate (0.348, 348 μg
`mm−3). Concentration-dependent toxicity. In this assay, mortality in
`both concentrations was observed to be significantly higher than
`the control (Fig. 14). Treatment with the higher dosage, 348 μg
`mm−3, resulted in 100% mortality upon the first day. The mor-
`tality of the lower dosage group, 0.348 μg mm−3, was observed
`to become significantly higher on day three and continued to
`increase until reaching 28% on day 14. The crystal structure of
`dicyclohexylammonium hydrogen phosphate is shown in Fig. 15.
`
`3.9 Concentration–response feeding assays
`Monocyclohexylammonium dihydrogen phosphate (2.28, 22.8,
`228 μg mm−3). Concentration-dependent toxicity.
`In this assay
`
`Figure 12. Means (±SE) of C. formosanus mortality following treatment of
`food source cellulose filter paper with dicyclohexylammonium hydrogen
`phosphate in no-choice assays with 20 workers (n = 3). Collection group
`J was used and the data were recorded daily for 14 days. The F value for
`percent mortality is 7.334. P and df values are 0.0003 and 3, 56.
`
`Neat cyclohexylamine (1.0, 1.7, 2.4 μg mm−3). Concentration-
`dependent toxicity. The three highest dosage groups displayed
`mortality significantly different from the control. Mortality for
`the highest dosage, 2.4 μg mm−3, became significantly different
`than the control group on day six and thereafter (Fig. 9). Sixteen
`percent mortality was observed on day six, after which it increased
`to 63% on day 14 (Fig. 9). Mortality for the second highest dosage,
`1.7 μg mm−3, became significant on day five and increased to
`41% on day 14 (Fig. 9). Mortality for the 1.7 μg mm−3 treatment
`became significant on day eight and increased to 33% on day 14
`(Fig. 9).
`
`3.6 Cellulose consumption
`Neat cyclohexylamine (0.626, 0.822, 1.253, 2.510 μg mm−3). Reduced
`In this assay, the three highest dosage
`cellulose consumption.
`groups, 0.822, 1.253 and 2.510 μg mm−3, resulted in consumption
`rates (per surviving termite) that were significantly lower than the
`control group (Fig. 10). The lowest dosage, 0.626 μg mm−3, did not
`result in a consumption rate per termite significantly different from
`the controls (Fig. 10).
`
`2044
`
`wileyonlinelibrary.com/journal/ps
`
`© 2017 Society of Chemical Industry
`
`Pest Manag Sci 2017; 73: 2039–2047
`
`Merck Exhibit 2203, Page 6
`Mylan Pharmaceuticals Inc. v. Merck Sharp & Dohme Corp.
`IPR2020-00040
`
`

`

`Cyclohexylamine induces antennae loss in Formosan termites
`
`www.soci.org
`
`Figure 13. Means (±SE) of C. formosanus mortality following treatment of food source cellulose filter paper with dicyclohexylammonium hydrogen
`phosphate in no-choice assays with 20 workers (n = 3). Collection group J was used and the data were recorded daily for 14 days. (A) Concentration was
`4.90 μg mm−3; (B) concentration was 9.80 μg mm−3; (C) concentration was 1.22 μg mm−3; (D) concentration was 2.45 μg mm−3. The F value for percent
`mortality is 9.07. P and df values are <0.0001 and 4, 70.
`
`2045
`
`Figure 15. Crystal structure of dicyclohexylammonium hydrogen phos-
`phate.
`
`Figure 14. Means (±SE) of C. formosanus mortality following treatment of
`food source cellulose filter paper with dicyclohexylammonium hydrogen
`phosphate in no-choice assays with 20 workers (n = 3). Collection group
`J was used and the data were recorded daily for 14 days. The F value for
`percent mortality is 139. P and df values are <0.0001 and 2, 42.
`
`Pest Manag Sci 2017; 73: 2039–2047
`
`© 2017 Society of Chemical Industry
`
`wileyonlinelibrary.com/journal/ps
`
`Merck Exhibit 2203, Page 7
`Mylan Pharmaceuticals Inc. v. Merck Sharp & Dohme Corp.
`IPR2020-00040
`
`

`

`www.soci.org
`
`B Grimball et al.
`
`Figure 16. Means (±SE) of C. formosanus mortality following treatment of food source cellulose filter paper with monocyclohexylammonium dihydrogen
`phosphate in no-choice assays with 20 workers (n = 3). Collection group J was used and the data were recorded daily for 14 days. (A) Concentration was
`2.28 μg mm−3; (B) concentration was 22.8 μg mm−3; (C) concentration was 228 μg mm−3. The F value for percent mortality is 6.424. P and df values are
`0.0008 and 3, 56.
`
`only mortality in the highest dosage group, 228 μg mm−3, was
`observed to be significantly different from the control (Fig. 16).
`Mortality became significantly different on day 10, reaching a
`value of 38%, and eventually increased to 100% on day 14
`(Fig. 16). Mortality for the two lower dosage groups was not sig-
`nificantly different from the control (Fig. 16). The crystal structure
`of monocyclohexylammonium dihydrogen phosphate is shown
`in Fig. 17.
`
`4 DISCUSSION
`Cyclohexylamine was found to be a toxicant to C. formosanus, as a
`neat compound, in sugar phosphate salt forms and in two different
`phosphate salt forms. The discovery was accidental and occurred
`during experiments screening rare sugars for termite toxicity,7
`where one of the sugars tested was I2P as the dicyclohexylam-
`monium salt. The toxicity manifested in concentrations at the
`LD30 –LD70 level as a most unusual specific sequential ablation of
`the termite antenna pseudosegments over a 3–14 day time frame.
`
`The mechanistic reason for specific loss of antennae is not known;
`however, the effect on termite workers is to cease normal behav-
`iors (feeding, moving in a two-dimensional plane, grooming; data
`not shown), which are stimulated and monitored by the antenna
`chemosensorium of the termite culture. Termites become increas-
`ingly moribund as their antennae are incrementally degraded,
`finally ceasing feeding and movement when the antennae are
`completely gone. Practically, the I2P and glycerol-3-phosphate
`salts of cyclohexylamine are too expensive to consider for termite
`toxicants, and neat cyclohexylamine is volatile, and noxious.
`Therefore we prepared the inexpensive mono- and dicyclo-
`hexylammonium hydrogen phosphate salts, made simply from
`phosphoric acid and cyclohexylamine, producing easily crystal-
`lized compounds that can be added to paper baits to effect a
`slow-acting toxin at 14 day LD30 –LD70 levels that have the poten-
`tial of seriously disrupting a colony with trophyllaxis feeding of
`the soldiers and queen (not tested in this study). It will be inter-
`esting to determine the mechanism of cyclohexylamine toxicity
`and antenna loss.
`
`wileyonlinelibrary.com/journal/ps
`
`© 2017 Society of Chemical Industry
`
`Pest Manag Sci 2017; 73: 2039–2047
`
`2046
`
`Merck Exhibit 2203, Page 8
`Mylan Pharmaceuticals Inc. v. Merck Sharp & Dohme Corp.
`IPR2020-00040
`
`

`

`Cyclohexylamine induces antennae loss in Formosan termites
`
`www.soci.org
`
`Rhinotermitidae), established in living urban trees in New Orleans,
`Louisiana, USA,
`in 3rd International Conference on Urban Pests.
`Graficke zavocy, Hronov, Czech Republic (1999).
`3 Messenger MT, Scheffrahn RT and Su NY, First report of Incisiter-
`mes minor (Isoptera: Kalotermitidae) in Louisiana, Fla. Entomology
`83:92–93 (2000).
`4 Lax AR and Osbrink WLA, United States Department of Agriculture –
`Agriculture Research Service research on targeted management
`of the Formosan subterranean termite Coptotermes formosanus
`Shiraki (Isoptera: Rhinotermitidae). Pest Manage Sci 59:788–800
`(2003).
`5 Wyss M, Brugger R, Kronenberger A, Rémy R, Fimbel R, Oesterhelt G
`et al., Biochemical characterization of fungal phytases (myo-inositol
`hexakisphosphate phosphohydrolases): catalytic properties. Appl
`Environ Microbiol 65:367–373 (1999).
`6 Veillon L, Muniruzzaman S, Henderson G and Laine RA, Toxic effects
`of 2-deoxy-d-galactose on Coptotermes formosanus (Isoptera:
`Rhinotermitidae) and symbionts. J Econ Entomol 103:1647–1656
`(2010).
`7 Veillon L, The biological activity of rare carbohydrates and cyclitols
`in Coptotermes formosanus. PhD thesis, Louisiana State University
`(2011).
`8 Smith WR, Amburgey TL, Henderson G and Ring DR, Facility for
`conducting field tests on Coptotermes formosanus at Louisiana State
`University Agricultural Center. For Prod J 54:26–28 (2004).
`9 Sheldrick GM, A short history of SHELX. Acta Crystallogr A64:112–122
`(2008).
`10 Farrugia LJ, WinGX and ORTEP for Windows (2012). Available:
`https://www.chem.gla.ac.uk/∼louis/software/downloads/farrugia_
`2012g.pdf
`11 SAS Institute, SAS/STAT User’s Guide Computer Program, version 6. SAS
`Institute, Cary, NC (1989).
`12 Bansemer S and Scheel T, Image Analyzer computer program, version
`3.6.4, Rostock, Germany (2010).
`13 Mannesmann R, Relationship between different wood species as a
`termite food source and the reproduction rate of termite symbionts.
`Z Angew Entomol 72:116–128 (1972).
`14 Maistrello L, Henderson G and Laine RA, Comparative effects of vetiver
`oil, nootkatone and disodium octaborate tetrahydrate on Coptoter-
`mes formosanus and its symbiotic fauna. Pest Manag Sci 59:58–68
`(2002).
`
`Figure 17. Crystal structure of monocyclohexylammonium dihydrogen
`phosphate.
`
`ACKNOWLEDGEMENTS
`The authors thank Professor Brian Marx for helpful discussions on
`statistical methods, and Professor Gregg Henderson for advice and
`information about Formosan termite behavior, providing termites
`from collections, and helpful suggestions for experiments.
`
`REFERENCES
`1 Grace JK, Ewart DM and Tome C, Termite resistance of wood species
`grown in Hawaii. For Prod J 46:57–60 (1996).
`2 Osbrink WLA, Woodson WD and Lax AR, Populations of For-
`mosan subterranean termite, Coptotermes formosanus (Isoptera:
`
`2047
`
`Pest Manag Sci 2017; 73: 2039–2047
`
`© 2017 Society of Chemical Industry
`
`wileyonlinelibrary.com/journal/ps
`
`Merck Exhibit 2203, Page 9
`Mylan Pharmaceuticals Inc. v. Merck Sharp & Dohme Corp.
`IPR2020-00040
`
`