BACTERIAL HYDROCARBON OXIDATION I. OXIDATION OF n-HEXADECANE BY A GRAM-NEGATIVE COCCuS' JAMES E. STEWART,2 3R. E. KALLIO, D. P. STEVENSON, A. C. JONES, AND D. 0. SCHISSLER Department of Bacteriology, State University of Iowa, Iowa City, Iowa, and Shell Development Company, Emeryville, California Received for publication March 17, 1959 The pioneering work of Sohngen (1913) es- (Brown and Strawinski, 1958; Dworkin and tablished that a variety of aerobic microorganisms Foster, 1956). Many reports in the literature are capable of growing at the sole expense of indicate that fatty acids, , or diverse paraffins. These findings have been have been recovered from cultures oxidizing extended and corroborated so often since then aliphatic paraffins (see Beerstecher, 1954, for that detailed documentation does not seem neces- complete citations). In no case, however, has a sary here, especially since the subject of microbial (or other intermediate) been identi- hydrocarbon oxidation has been comprehensively fied which has a carbon skeleton identical to the and frequently reviewed (Beerstecher, 1954; alkane being oxidized, thus leaving in question Davis and Updegraff, 1954; ZoBell, 1950). the locus of the primary enzymatic attack on the Despite the extensive literature on the subject alkane molecule. of the mechanisms by which bacteria and other This paper summarizes preliminary findings of microorganisms attack hydrocarbons, the re- a study undertaken to investigate the mecha- actions have not been subjected to intensive nisms of microbial attack on aliphatic paraffins. experimental study. For example, an important goal in the establishment of a metabolic mecha- MATERIALS AND METHODS nism, namely, the isolation and identification A gram-negative coccus (designated as H.O.1 of an intermediate compound reasonably close during this study), with ability to utilize a wide to the substrate in terms of structure, has been variety of organic compounds for growth, was accomplished in the microbial oxidation of ali- isolated from hexadecane-mineral salts enrich- phatic hydrocarbons only three times; of these ments. The enrichments were incubated at room only one is alkane oxidation. Birch-Hirschfeld temperature on a gyratory shaker and after 7 (1932) detected the presence of acetaldehyde in days a transfer was made to fresh medium, fol- cultures of Mycobacterium lacticola growing at lowing which 6 serial transfers at 48-hr intervals the expense of acetylene. Bruyn (1954) isolated were made to flasks containing fresh mineral- and identified 1, 2-hexadecanediol in cultures of hexadecane medium. Pure cultures were ob- Candida lipolytica growing in a mineral salts- tained by streaking on mineral agar and in- hexadecene-1 medium. Finally, chemical evi- cubating in a paraffin atmosphere obtained by dence has been reported which strongly suggests adding 1.0 ml of sterile hexadecane to a circle of that methanol, formaldehyde, and formic acid filter paper previously inserted in the lid of the are produced from methane by Methanomonas petri plate. The detailed microbiological at- methanooxidans and Pseudomonas methanica tributes of all isolates from hexadecane as well as other hydrocarbon enrichments will be the ' This study was supported by a grant from the subject of a separate report. All isolates thus far Petroleum Research Fund administered by the have been gram-negative cocci. Strain H.O.1 was American Chemical Society. Grateful acknowl- chosen for study because of its profuse and edgement is hereby made to the donors of said homogenous growth on alkanes and its relatively fund. low rate. The 2 American Chemical Society, Petroleum Re- endogenous respiratory organism search Fund Predoctoral Fellow. is catalase positive and does not exhibit 3Present address: Technological Station, Fish- globules at any stage of growth. eries Research Board of Canada, Halifax, Nova Large scale cultures were set up with 1 L of Scotia. mineral medium in a 3-L Fernbach flask inocu- 441 442 STEWART, KALLIO, STEVENSON, JONES, AND SCHISSLER [VOL. 78 lated with 1 pexr cent of 48-hr growth of the or- gen analysis showed C = 79.48 per cent; H ganism on the s,,ame medium. Sterile hexadecane 13.28 per cent; and 0 = 7.24 per cent (by dif- was added at tble time of inoculation to provide ference). Phenolphthalein titration in ethanol a carbon source of 1 per cent (v/v). Cultures indicated that the substance was not an acid. were grown on Ea gyratory shaker at 250 rpm at The infrared spectrum of a carbon tetachloride room temperatu]re. soluti6n of the material showed ibsorption bands Titratable aciLdity, pH, and consump- at 3.42 and 3.51 p, characteristic of the C-H tion were determined by standard methods. stretch motion of -CH2-- or -OH. groups, Growth was diifficult to measure during the and at 6.83 ps characteristic of -CH,- bending course of an experiment because of the forma- motions. Additionally the material showed bands tion of stable emulsions of the hydrocarbon under at 5.74 p characteristic of carbonyl groups and the combined action of bacterial attack and con- another band at 8.53 p. The absorptivity of the stant agitation. 'The method finally adopted con- 5.74 p band was found to be 89 L per 100 g cm. sisted of transfeirring 6.5 ml of culture fluid to a In isopropyl , this band separated into pyrex centrifuge tube having a narrow tip gradu- two components of about equal intensity at 5.73 ated in hundredlths from 0.1 to 0.4 ml. Tubes p and 6.79 , a unique diignostic crcistic were centrifuged[ for 30 min at 1500 X G and of the structure (A. C. Jons, bed growth was reciorded as milliliters of cells per oem,. Taking the molal absorptivity of 6.5 ml. the 5.t4 band of (im CCl) to ble 450 L A WestinghouLse type LV mass spectrometer, per mol cm, the absorbance of the bacterial modified to have a nominal resolving power of oxidation intermediate corresponds to that of 1 in 350, with a heated inlet system was used to an ester with molecular weight equal to 500 t 50. record the mass spectra. The mass spectra were The nzwss spectrum showed a stront ion intensity scanned with coinstant ion accelerating potential at a position corresponding to a C, hydrocarbon by varying the mass analyzing field. The ion indicating the molecular weight of the material source temperatiure was about 200 C and 75 v to be.480 ± 14. The next strong intensities on ionizing electron:s were employed to generate the a descending mass scale were a set of three peaks ions. of mass to charge ratio 258, 257, and 256, and Other procedures used are outlined in Results a set of two at 225 and 224. In thse iets, those and Discussion. at 257 and 224 were much more intee.f16ma to charge ratios of these ions were Witively RESUFLT5 AND DISCUSSION identifid by the addition of an authentic sample Preliminary cibservation on earlier isolates of normal octadecane and recording of the mass szimilarj;1l1ll1al toUV TI(-0Ls.L.N111ALMAIUUIUCUVUhadl indioaf.wlUWCVthat.llULmlA1inDnir rL-vnro+AhblVWl an10t.inlm on hexadecane a marked drop in pH occurred, The intensities of the ions 258 and 225 relative the lowest values occurring about 48 hr after to those of 257 and 224, respectively, showed that inoculation. While the nature of the change in both the 257 and 224 ions contained 16 ti pH was unknown it seemed logical to assume that carbon atoms. Hence, the most probable empiri- some intermediate product of an acidic nature cal formulae of these ions would be C1HJO62+ was accumulating in the culture fluid. Accord- (257) -nad CuHn32+ (224). These data suggest ingly, 1 L of mineral-hexadecane medium was that the substance in question- wa either inoculated with one isolate (strain F IV) and CQi%HS OCI.H,, cetyl pdlmitat;a *di -of after 48 hr of incubation the cells were harvested C1..0, or possibly C.H1^(DH(OH3)O tie and the culture filtrate was evaporated under hemiacetal of cetyl and t l reduced pressure to a volume of approximately Hoeb,the unambiutinra& ee 200 ml. The concentrated filtrate was dissolved for an* hse combined with the.ma-e 64 - in an excess of boiling ethanol, filtered, and con- metri ati cle tt te centrated. White crystals formed upon cooling; is cetyl palmitate. Incidentaly, the these were removed by filtration and recrystal- truinsowed virtually no oin the" w e lized from ethanol to a constant melting point 270 to 470, indicating that both the Alyigoup of 46 to 48 C. Elemental analysis showed that of the eisterae nomal. nitrogen and sulfur were absent. Carbon-hydro- In later identification of oters the.m sale 1959] BACTERIAL HYDROCARBON OXIDATION4443 was calibrated by the addition of an authentic Ester Forriotion Growth aLMoles /IOOML ML/65ML Titrotable sample of n-C36H74 which resulted in the reduction 55C- 0.5 of the uncertainty in the determination of the molecular weight of the intermediate from -14 440 - 04 ( mass units to ±+'y mass units. This alkane, I n-C36H74 gives a set of fragment ions of formula 330 0.3

CnH2,+1+ and CnH2n+ 34 > n > 12, all of about 220 0.2- i equal intensity. An ester, CmnH2m+lCO2CkH2k+1, has its parent peak at 3 mass units greater than 110: 0.1 Cm+k+3H2(m+k+3)+l. It has been established that the characteristic fragment ions in an ester mass spectrum are CmH2m+lCO2H2+ and CmH2m+1- CO2H+, with relative intensity approximately Figure 1. Growth and ester production of the 5/1, and CkH2k+ where this ion is of intensity gram-negative coccus H.O.1 growing on n- those of the acid moiety. The hexadecane. Organisms were grown on the mineral intermediate to medium of McClung augmented with 2 per cent ester is, as noted, (n)CmH2m+1CO2(n)CkH2k+1. n-hexadecane (v/v). Growth was at room tem- Firm identification of this intermediate de- perature on a New Brunswick gyratory shaker rived from an alkane by bacterial oxidative action (250 rpm). Determinations were carried out as is strikingly significant since the carbon skeleton set forth under Materials and Methods. of the paraffin appears to be preserved at this level of oxidation. Under similar conditions, cetyl reached a maximal value at about 30 hr, some- palmitate was also isolated from cultures of the what prior to maximal ester production, and organism used for the remainder of this study thereafter lysis of cells occurred so that cell (H.O.1). The major significance of this identifica- harvest was essentially zero at 200 hr. Although tion lies in the fact that it permits an experi- no claim can be made for the precision of the mental approach to an analysis of the mechanisms growth measurements, the curve obtained is real involved in the process. Accordingly, an analysis inasmuch as subsequent experiments showed that of bacterial growth vs. ester production was it is possible to obtain a more conventional growth carried out. Growth was measured as noted under with other substrates (lactic acid, glycerol). Materials and Methods. The ester content was It may be argued on the basis of the data in determined on 6-ml samples of the growing cul- figure 1 that cetyl palmitate is a nonfunctional tures as follows. The sample was placed in a 50- end product accumulating from oxidized hexa- ml volumetric flask, 25 ml of 1:2 ether-ethanol decane, but tests indicated that the ester sup- mixture were added and boiled in a 90 C water ported slow growth of the test organism at al- bath. After cooling, the contents of the flask were kaline pH values. The organism also demon- brought to volume with the ether-ethanol mixture strated an ability to oxidize the compound and filtered. The supernatant fluid (41.2 ml) was slowly when tested by conventional manometry. transferred to a 200-ml round bottom flask and Both processes doubtless are profoundly affected evaporated to dryness in a flask evaporator. The by the insoluble and nonwettable nature of the nonvolatile ester concentration in the residue was wax. The reason(s) for the lysis of cells under determined following the directions of Bauer and these conditions has not yet been ascertained. Hirsch (1949) for hydroxamic acid assay of water A number of questions arose at this stage of insoluble esters. Tripalmitin and cetyl palmitate the study and foremost among these was whether standard curves. The were used to construct or not the ester actually was produced from from the acid color developed hydroxamic by hexadecane or from some contaminant. The pro- the addition of was read in the Beck- Fe(CI04)3 of man model DU spectrophotometer at 520 m,. hibitive cost very highly purified paraffin Data thus obtained are plotted in figure 1. samples (American Petroleum Institute Stand- Under the conditions of the experiment, ester ards) precluded the use of anything but com- production reached a maximum at about 70 hr mercial hexadecane in the relatively large scale and thereafter declined slightly. The growth growth experiments. The substrate hexadecane curve is particularly interesting inasmuch as it used (Humphrey-Wilkinson, Inc.) was subjected 444 STEWART, KALLIO, STEVENSON, JONES, AND SCHISSLER [VOL. 78

Ester Formation (Growth Titrotable ml, ,;qu,m Acidlty O26 = 96.7 i 0.1 per cent Moles/ IOOML ML lbV.ZIML ...... &...... 500r ESTER 016017 = 0.218 i 0.013 per cent 016018 = 3.04 i 0.05 per cent 400+ pH O28 = 0.03 i 0.01 per cent 6ROWTH 300+ 0.3± -.15 Preliminary experiments with ordinary oxygen pH TlTRA TAB LE indicated that increasing the oxygen tension over 200- 0.2 - AC/D/ TYz,t ,.___ - - 1.0 80 the growing cultures decreased the time required 100t 0.1 *PH 0.5 70 for ester formation and increased somewhat the initial growth rate of the organisms. Maximal (I 0 6 12 18 24 30 36 ester formation could be achieved in 24 hr when Time In Hours an atmosphere of pure oxygen was provided. Figure 2. Growth and ester production of the Some changes in the total crop of organisms were gram-negative coccus growing on n-hexadecane evident during growth under oxygen (figure 2) under pure oxygen. Conditions as in figure 1. but for technical reasons it was decided to per- form 018 incorporation experiments under oxy- to mass analysis, and apart from n-C16H34 the gen. A 4-L round bottom flask equipped with two following were found: stopcocks was charged with 1 L of the basal C17H36 (as normal) 0.5 per cent medium, 10 ml of hexadecane, and inoculated C16H32 1.2 per cent with 10 per cent of a liquid culture of the or- C15H32 (as normal) 1.1 per cent ganism. The flask was evacuated, flushed with nitrogen, evacuated again, and filled with the Since chemical analysis showed that under 018 atmosphere. The incubation was carried out the conditions of the experiment (figure 1) 2.5 g at 25 C and 018 was admitted as required. Twenty of ester (calculated as cetyl palmitate) were four hours later the ester was isolated and sub- formed from 7.73 g of hexadecane (d20 .7734) jected to mass analysis and compared with a by growing cultures, the amount of ester pro- sample of cetyl palmitate isolated in a con- duced exceeds considerably the impurity present. comitant experiment but under ordinary oxygen. An equally important question, apart from Following is an account of the analytical method. teleological considerations, was concerned with As has been pointed out the characteristic set the mechanism(s) which was by cetyl palmitate of ions formed from hexadecane. It seemed that the most of the ester mass spectrum is that with mass to charge ratio logical way the hydrocarbon structure could be equal to 259, 258, 257 and breached was through the formation of 1-hexa- 256. The primary peaks of this set are 257 = decylhydroperoxide. Two such C16H3302+ and 256 = C16H3202+, dissociation molecules might give rise to two molecules of fragments of the ester. The peaks of m/q = palmitaldehyde which by a mechanism analogous 259 and 258 are "isotope peaks" (259 to the Tishchenko reaction would lead to cetyl C16H33016018 + C14C'3HU02; 258 ' C15C'3H3302 palmitate. Mosher and Wurster (1955) have + C14C13H3202 + Cl6H32O16018). shown that the thermal decomposition of n-butyl The analytical procedure for incorporation of hydroperoxide leads to (among other things) atmospheric oxygen had as its primary data the n-butyl-n-butyrate; the ester represented 27.5 intensities of the four ions. From the relative per cent of the decomposition products. This intensity, 259/257, a correction to 258 for the chemical analogy was tempting to consider and C13 + 018 corresponding to the acid ion was ob- direct proof of the hypothesis was sought by per- tained. From the corrected of forming the bacterial intensity peak 258 ester-producing experi- the intensity of 257 could be corrected for the ments under an atmosphere of 018 enriched oxy- C13 from the acid ion: gen. Oxygen is directly incorporated into the hydroperoxide and the ester should therefore = (258) - show 100 per cent incorporation of 018. Electrol- (258c) (257) (256) ysis of D2018 (018 = 8.095 X normal) yielded 018 enriched oxygen of the following composition (257c) = (257) - (256) (basis oxygen): (257) 1959] BACTERIAL HYDROCARBON OXIDATION 445

TABLE 1 Mass analysis of cetyl palmitate samples produced by the gram-negative coccus from hexadecane under 08-enriched oxygen and ordinary oxygen

Mass to Charge (m/q) Ion Intensities Origin of Cetyl-Palmitate 259 258 257 256

Synthetic product ...... 0.0187 0.185 1.000 0.229 Coccus under 08 enriched oxygen ...... 0.0385 =1: 6* 0.192 i 2 1.000 0.253 4 2 Coccus under ordinary oxygen ...... 0.0192 9 0.186 4i 5 14100 0.241 4 6 * Uncertainties are standard deviations from the mean of 6 measurements. TABLE 2 The relative C12H330s6018 for the normal ester, Corrected m/q ion intensities of cetyl patmitate 0.0031 9, is in reasonable agreement with produced from hexadecane 0.0038 found in normal oxygen when the mag- nitude of the C13 correction (0.0170) is noted. Corrected m/q Ion Intensities Origin of Ester The 018 content of the cetyl palmitate produced 018 (258,) (257,) (259') (258') (257') under is 1.17 :1 0.1 per cent. The increased uncertainty of this final figure allows for the 018-Enriched 0. 182 0.954 0.04040.191 1.000 neglect of 017 and the natural abundance of oxygen deuterium in the calculations. To some extent Ordinary oxygen 0.181 0.95610.020110.190 1.000 there is partial account taken of these (017 and D) in the "effective C13,"' a = 0.0119. Thus, the a = 0.0119. ester produced under 02, containing 1.55 per Then relative cent of 018, contained 1.17 per cent of 018, an incorporation of approximately 75 per cent C"H33016018 + C"1C"3H330"8 = (259/257,) (1.17/1.55 = 0.754). and relative Rather than supporting the Tishchenko mecha- nism the 018 incorporation data differ from the C15C03H330"6 = (258,) /(257,). expected value significantly enough to preclude The small concentration if 017 was neglected. the hypothesis. The data obtained, however,, fit If a is the effective relative abundance of C13, the notion that two 1-hydroperoxide molecules then are produced directly from the paraffin and these relative C12CI3H33016 = 16a are reduced to the normal alcohol. One alcohol relative C12C13H.016 = 120a2 molecule is oxidized to (in the case of hexadecane) palmitic acid and during the process the 018 and relative C12H33016018 originally present in the alcohol is randomized (259) 120 (258 ) l2 with 016 from water of the medium. The hydro- (257c) 256 257C peroxide -+ normal alcohol mechanism is a tenta- tive suggestion since several other hypothetical In addition to the two samples of cetyl pal- mechanisms are also consistent with the data mitate produced from hexadecane by the coccus but a choice among these cannot be made at this under 018 and ordinary oxygen, a synthetic time. Experiments in progress are expected to sample of cetyl palmitate prepared by Dr. C. D. shed additional light on the problem. The pro- Wagner by reaction of excess Cj5H31COCl with posed mechanism (figure 3) is predicated on the C16H33OH was examined in the same way. The assumption that no great isotope effect obtains results are shown in table 1. Corrected m/q ion in the system with regard to the utilization of intensities are shown in table 2. From these data oxygen by the organism. the following may be obtained: Additional evidence for the early participation relative C16H33O560O of hexadecanol-1 in cetyl palmitate formation '0l8 ester" 0.0234 6 is found in figure 4 wherein ester production dur- 2Ol6 ester" 0.0031 i 9 ing growth of the organism on hexadecanol is 446 STEWART, KALLIO, STEVENSON, JONES, AND SCHISSLER [VOL. 78

2 CH3(CH2),4CH3 + 202 oxidizing systems as being analogous to lipox- idase (reviewed by Mason, 1957), which forms at the methylene bridges of 2CH3(CH2),ICH20'8018H methylene-interrupted multiply unsaturated fatty acids. However, there is no evidence for 2CH3(CH2)ICH2018H the formation of olefins during the bacterial I oxidation of alkanes. On the contrary, fragmen- H 18- CH3(CH2),4 C = 0 I tary data from this laboratory and reports from +H2016 8 other laboratories (Bruyn, 1954) strongly sug- gest that olefin oxidation proceeds in a different CH3(CH2)14CI H manner from paraffin oxidation, i. e., via forma- -016 tion of a-glycols at the double bond. -2H jH Implication of the postulated 1- hydro- LCH3(OH2)14Cz:IGj0H CH3(CH2),4CH20'8H peroxide is admittedly based on indirect evidence. (50%5 confrbut/ion of 0/8) The formation of hydroperoxides in nonbiological (018contribuh=on=° =25%)JJ oxidation of paraffins is well known (Vaughn Rondomlzolion of 0I%,n ocid w/th 06fromHr0 and Rust, 1955) and it should be noted that in nonbiological paraffin oxidations at low temper- CH3(CH2),4- C- o08 (CH2) SCH3 atures there is no evidence suggesting the inter- 11 18,16 mediate formation of olefins. Imelik (1948a, b) (IB incorporation = 75%) noted the formation of "peroxides" during the oxidation of cyclohexane by growing cultures Figure S. Postulated reaction sequence for of Pseudomonas aeruginosa. "Peroxides" (Eisen- formation of cetyl palmitate from n-hexadecane berg, 1943) have been detected during the growth bv the coccus (H.O.1). of the gram-negative coccus on hexadecane but only when metallic ions, especially iron, were compared with ester produced at the expense of kept to extremely low concentrations in the growth on palmitic acid. Clearly, ester production growth medium. At present it is impossible to by the organisms from hexadecanol is far in ex- test directly the suggestion that 1-hexadecylhy- cess of ester produced from palmitic acid. Maxi- droperoxide is an intermediate in the oxidation mal ester accumulation is 650 ,moles per 100 of hexadecane because 1-hexadecylhydroperoxide former case as ml in the compared with 30,umoles is not known. A number of the lower 1-hydro- of ester per 100 ml in the latter. This would be the expected result if palmitic acid were pro- peroxides have been prepared, however (Williams duced from hexadecanol followed by a direct and Mosher, 1954), and through the courtesy of esterification of the acid and the alcohol. The ester Ester Formation produced by the organism from hexadecanol was ,A Moles / 100 Ml isolated and by the techniques described above was shown to be cetyl palmitate. Palmitaldehyde as a growth substrate yielded inconclusive evi- dence and the results appeared to be confused by side reactions and condensation products of the aldehyde (other than ester). It is not known at present whether the first step in bacterial paraffin oxidation is an oxygen transferase or a mixed function oxidase (Mason, 1957). No cell-free paraffin oxidations have been observed despite considerable effort to develop 70 105 14 such a system. In fact, thus far, only growing Time in Hours cells have been found to oxidize paraffins at rates Figure 4. Comparison of ester formation by the sufficiently rapid to result in ester accumulation. coccus (H.O.1) growing on hexadecanol and on The formation of hydroperoxides by lipoxidase palmitic acid as sources of carbon. Conditions as action makes it tempting to consider the alkane in figure 1. 1959] BACTERIAL HYDROCARBON OXIDATION 447

Dr. Harry MAosher of Stanford University a small sample of 1-decy1hydroperoxide was ma(le avail- able for testing. Cells grown at the expense of n-decane or hexadecane readily oxidized the 1-hy- (Iroperoxide as illustratedl in figure 5. The 1-dee- ylhydroperoxide was found to be 90 per cent pure by infrare(d analysis (H. S. MIosher, personal conmnunication). The nature of the contaminating substances is not known but on the basis that 10 per cent of the dlecylhydroperoxide sample consists of the corresponding aldehyde, a not uncommon re(luction product of hydroperoxides, the 2.05 of apparent ,umoles hydroperoxide would 20 40 60 80 100 have containe(l approximately 0.20 ,umoles of Minutes the aldehy(le. On the assumption that total Figure 5. Oxidation of 1-decylhydroperoxide by oxidation oc(urs, i. e., no oxi(lative assimilation, cells of the gram-negative coccus H.O.1. Grown this amount of decvlaldehvde would account for at the expense of n-decane. Phosphate buffer the corsumption of 7.1 Amoles of oxygen (159 (0.05 M, pH 7.2); 30 C; 4 mg dry weight of cells AL). This figure added to the endogenous value per vessel. The hydroperoxide, added as indicated, obtained at 90 min (110 ,uL) equals 269 uL, which was emuilsified in distilled water juist, pr-ior to use. subtracted from the 580 ,uL obtained for the Decane was API Standard emulsified in water hydroperoxide at an apparent 2.05 j,mole level under N2. indicates that 311 ,IL of oxygen were accounted for by hydroperoxide itself, considerably in ex- with nonbiological paraffin oxi(lations. Pope cess of the total as well as the rate observed for et al. (1929a, b) claimed that autoxidation (per- decane. The organism oxidizes (lecyl alcohol and oxide formation) of paraffins occurred on the (lecanoic acid as well as decylaldehyde so that methyl group at the end of the largest alkyl essentially the same considerations apply should chain, but this hypothesis has now been replaced the impurity in the peroxide be the acid, the by the notion that the hydrogen atoms in a hydro- alcohol, the aldehyde, or any combination of carbon are all vulnerable to the free radical attack these compounds. The fact that alkane-grown which is the accepted mechanism for paraffin cells rapidly oxi(lize 1-d(ecylhvdroperoxi(le as well autoxidations (Walling, 1957). In a recent re- as the paraffins may, of course, be a coincidental view (Vaughn and Rust, 1955) on the subject, relationship. However, it does suggest that the it has been pointed out that the frequency of 1-hydroperoxide postulation is reasonable. Un- attack at any one position will be complicated fortunately, the apparenit toxicity of the hydro- by and depend upon such faetors as hydrogen peroxi(le at the higher concentrations used pre- atom reactivity, the numbers of available atoms cluded the isolation of intermediate compounds at any position, and possibly steric factors. Gen- from reaction mixtures. Efforts are being made to erally, the order of increasing reactivity of hydro- synthesize the 1-hydroperoxide of hexadecane gen atoms will be primary, secondary, and ter- and to construct experiments which will more tiary; thus in the case of normal paraffins the directly establish the participation of 1-alkyl expected attack and consequent hydroperoxide hvdroperoxides in the bacterial oxidation of formation should occur among the methylene paraffins. earbons more or less randomly. This concept is The (lata recorded here certainly indicate that supported by the findings of Benton and Wirth bacteria attack the terminal or 1-position of (1953) who found that all four pairs of methylene hexadecane. Information obtained with 018 en- groups in n-decane were equally vulnerable to riche(d oxygen show that oxygen is an obligatory oxidative attack whereas the terminal methyl reactant in the preliminary bacterial oxidative groups were largely untouched by free radieal step of hydrocarbons as has been previously hydrogen abstraction. The nonspecific attack on suggested (Imelik, 1948a, b; Hanson and Kallio, the methylene groups is similar to free radical 1957). The terminal locus of biological oxidation chlorinations but is not in accord with the find- on alkane molecules is interesting to compare ings of Ivanov et al. (1948) who found that the 448 STEWART, KALLIO, STEVENSON, JONES, AND SCHISSLER [VOL. 78

2-carbon position is the favored position for DAVIS, J. B. AND UPDEGRAFF, D. M. 1954 Mi- hydroperoxide formation in alkane autoxidations. crobiology in the petroleum industry. Bac- In any case, however, there is general agreement teriol. Revs., 18, 215-238. that terminal methyl groups of paraffins are the DWORKIN, M. AND FOSTER, J. W. 1956 Studies least preferential locus for nonbiological oxida- on Pseudomonas methanica (Sohngen) nov. comb. J. Bacteriol., 72, 646-659. tive attack. EISENBERG, G. M. 1943 Colorimetric determi- A prior report of 2-hexadecanone as an inter- nation of hydrogen peroxide. Ind. Eng. mediate in bacterial paraffin oxidation was found Chem. Anal. Ed., 15, 327-328. to be in error (Stewart and Kallio, 1957). The HANSEN, R. W. AND KALLIO, R E. 1957 In- 2-hexadecanone apparently arises as an artefact ability of nitrate to serve as a terminal when hexadecane is vigorously emulsified in an oxidant for hydrocarbons. Science, 125, 1198- air atmosphere-a technique employed in earlier 1199. studies. Subsequent tests have shown that in IMELIK, B. 1948a La croissance de Pseudomonas systems devoid of organisms small amounts of aeruginosa sur les petroles. Compt. rend., iodoform yielding substances are produced from 226, 1227-1228. IMELIK, B. 1948b Oxydation du cyclohexane par paraffins by such emulsification procedures. Pseudomonas aeruginosa. Compt. rend., 226, 2082-2083. SUMMARY IVANOV, K. I., SAVINOVA, V. K., AND ZHAKHOV- A gram-negative coccus isolated from hexa- SKAYA, V. P. 1948 Peroxide compounds of decane enrichment cultures grew profusely on 2,7-dimethyloctane. Doklady Akad. Nauk S. S. S. R., 59, 703-706. n-hexadecane and accumulated a substance identi- MASON, H. S. 1957 Mechanisms of oxygen fied by infrared spectroscopy and mass spectrom- metabolism. Advances in Enzymol., 19, 79- etry as cetyl palmitate. Atmospheric oxygen 233. participated directly in the oxidation process as MOSHER, H. S. AND WURSTER, C. F. 1955 De- demonstrated by 018 incorporation into cetyl composition of primary hydroperoxides. J. palmitate to the extent of 75 per cent. The data Am. Chem. Soc., 77, 5451. were interpreted to indicate formation of cetyl POPE, J. C., DYKSTRA, F. J., AND EDGAR, G. palmitate from the esterification of cetyl alcohol 1929a The vapor phase oxidation of isomeric and palmitic acid produced from 1-hexadecylhy- octanes. I. Normal octane. J. Am. Chem. droperoxide which was suggested as the first Soc., 51, 1875-1889. intermediate in the bacterial of n- POPE, J. C., DYKSTRA, F. J., AND EDGAR, G. 1929b oxidation The vapor phase oxidation of isomeric octanes. hexadecane. II. Octanes with branched chains. J. Am. Chem. Soc., 51, 2203-2213. REFERENCES S6HNGEN, N. L. 1913 Benzin, Petroleum, Pa- BAUER, F. C., JR., AND HIRSCH, E. F. 1949 A raffinol und Paraffin als Kohlenstoff-und new method for the colorimetric determination Energiequell fuer Mikroben. Zentr. Bak- of the total esterified fatty acids in human teriol. Parasitenk., Abt. II, 37, 595-609. sera. Arch. Biochem., 20, 242-250. STEWART, J. E. AND KALLIO, R. E. 1957 An BEERSTECHER, E., JR. 1954 Petroleum microbi. intermediate in the bacterial oxidation of ology, 1st ed. Elsevier Press, Inc., Houstoni, of n-hexadecane. Bacteriol. Proc., 1957, 134. Texas. VAUGHN, W. E. AND RUST, F. F. 1955 Low BENTON, J. L. AND WIRTH, M. M. 1953 PositioIn temperature oxidation of paraffin hydrocar- of radical attack during oxidation of long bons; oxidation of paraffin wax. In The chain paraffins. Nature, 171, 269. chemistry of petroleum hydrocarbons, Vol. II, BIRCH-HIRSCHFELD, L. 1932 Die Umsetzung von pp. 309-323. Edited by B. T. Brooks, C. E. Acetylen durch Mycobacterium lacticola. Boord, S. S. Kurtz, Jr., and L. Schmerling. Zentr. Bakteriol. Parasitenk., Abt. II, 86, Reinhold Publishing Corp., New York. 113-129. WALLING, C. 1957 Free radicals in solution. BROWN, L. R. AND STRAWINSKY, R. J. 1958 John Wiley and Sons, Inc., New York. Intermediates in the oxidation of methane. WILLIAMS, H. R. AND MOSHER, H. S. 1954 Bacteriol. Proc., 1958, 122-123. Peroxides. I. n-Alkyl hydroperoxides. J. BRUYN, J. 1954 An intermediate product in Am. Chem. Soc., 76, 2984-2987. the oxidation of hexadecene-1 by Candida ZOBELL, C. E. 1950 Assimilation of hydro- lipolytica. Koninkl. Ned. Acad. Wetenschap. carbons by microorganisms. Advances in Proc. Ser. C, 57, 41-45. Enzymol., 10, 443-486.