Vol. 59, No. 2APPLIED AND ENVIRONMENTAL MICROBIOLOGY, Feb. 1993, p. 365-3720099-2240/93/020365-08$02.00/0Copyright ©D 1993, American Society for Microbiology

A Strategy for Rotation of Different Bacteriophage Defensesin a Lactococcal Single-Strain Starter Culture Systemt

WESLEY D. SING' AND TODD R. KLAENHAMMERl 2*

Departments of Microbiology' and Food Science,2 Southeast Dairy Foods Research Center,North Carolina State University, Raleigh, North Carolina 27695-7624

Received 3 September 1992/Accepted 16 November 1992

A new strategy for starter culture rotations was developed for a series of phage-resistant clones geneticallyderived from a single strain of Lactococcus lactis subsp. lactis. Phage-resistant derivatives carrying differentdefense systems were constructed via conjugation with various plasmids encoding abortive infection (Abi/Hsp)and/or restriction and modification (R/M) systems of different specificity. The plasmids included pTR2030(Hsp+ R+/M+), pTN20 (Abi+ R+/M+), pTRK11 (R+/M+), and pTRK68 (R+/M+). Selected phage-resistanttransconjugants or transformants were evaluated in different rotation sequences through cycles of theHeap-Lawrence starter culture activity test in milk contaminated with phage and whey from the previous cycle.When used in consecutive sequence, derivative strains carrying the R/M systems encoded by pTN20, pTRK11,and pTRK68 retarded phage development when the initial levels of phage contamination were below 102PFU/ml but not when levels were increased to 103 PFU/ml. Use of a derivative bearing pTR2030 (Hsp+ R+/M+)at the beginning of the rotation prevented phage development, even when the initial levels of phagecontamination were high (106 PFU/ml). Alternating the type and specificity of R/M and Abi defenses throughthe rotation prevented phage proliferation and in some cases eliminated contaminating phages. A modelrotation sequence for the phage defense rotation strategy was developed and performed successfully over ninecycles of the Heap-Lawrence starter culture activity test in the presence of high-titer commercial phagecomposites. This phage defense rotation strategy is designed to protect a highly specialized Lactococcus strainfrom phage attack during continuous and extended use in the dairy industry.

Studies on phage-insensitive strains of lactococci haveuncovered the genetic basis for bacteriophage resistance (6,26, 31, 32-34, 46, 48). Currently, three types of natural phagedefense mechanisms are known and include prevention ofphage adsorption (Ads) (8, 51), restriction and modification(R/M) (3, 4, 9, 10, 15, 19, 50, 59), and abortive infection (Abi;also designated Hsp or Rbs in some published reports [2, 6,7, 9, 10, 12, 20, 25, 27, 35, 38, 44, 46]). All three classes ofresistance are commonly encoded by plasmid DNA ele-ments. In some cases, the genes encoding R/M and Abi existon the same plasmid, providing complementary defenses (6,9, 16, 32, 34). In other cases, multiple R/M systems existwithin a single strain and provide elevated levels of phagerestriction (3, 4, 10, 29).The development of gene transfer systems in lactococci

(37, 45) has provided opportunities to construct strains withimproved characteristics. The existence of conjugal transfer(Tra+) and phage resistance (Abi or Hsp or R/M) determi-nants on plasmids such as pNP40 (Tra+ Abi+), pTR2030(Tra+ Hsp+ R+/M+), pTN20 (Tra+ R+/M+), pCI1750 (Tra+Abi+), and pAJ1106 (Tra+ Hsp+) facilitates the constructionof phage-resistant starter strains by using simple conjugalstrategies (6, 15, 19, 27, 28, 44, 46, 54, 58). Transconjugantscarrying pTR2030 (Hsp+ R+/M+) have survived prolongeduse in milk fermentations (48, 52). Phages resistant to theHsp and R/M mechanisms have now been detected (1, 17,18) and thus illustrate the inevitable appearance of newphage upon the prolonged use of any single defense (orcombination of defenses) within industrial starter strains.

* Corresponding author.t Paper FSR92-30 of the Journal Series of Department of Food

Science, North Carolina State University, Raleigh, NC 27695-7624.

Traditional starter culture programs have depended on therotation of numerous lactococcal strains to minimize failuresdue to bacteriophage attack (21, 31, 42). When lytic phageappear in whey samples during a defined starter culturerotation program, new strains which are ideally unrelatedand not attacked by the same phage species or strains areincorporated into the rotation scheme. In many cases, thelongevity and starter activity of complex strain rotations areunpredictable and can lead to early failure, especially whena phage-related strain is unknowingly added to the rotationsequence (40, 61). Prolonged rotations involving numerousstrains increase the level and diversity of phages contami-nating the plant (13, 61), thereby also increasing the geneticpotential through which new phages may emerge. Multiple-strain starter culture programs, which rely on composites offive to six phage-unrelated strains, have been very success-ful because fewer strains are used and this in turn limitsphage diversity and population levels in the plant (39, 41,61).With the ability to genetically construct phage-resistant

lactococci, strategies can now be based on differences be-tween defined mechanisms. In this study, we have devel-oped and evaluated a novel approach for culture rotation byusing isogenic derivatives from a single strain which harbordifferent phage defense mechanisms. Alternating the type orspecificity of mechanisms was used to protect one bacterialhost strain by suppressing the development of new phageswhich may be resistant to any single mechanism or group ofmechanisms.

MATERIALS AND METHODS

Bacterial strains, plasmids, and bacteriophages. The bacte-rial strains, plasmids, and bacteriophages used in this study

365

on May 15, 2020 by guest

http://aem.asm

.org/D

ownloaded from

366 SING AND KLAENHAMMER

TABLE 1. Bacteria, plasmids, and phages

Strain, phage, Relevant characteristicsa Source or referenceor plasmid

L. lactis subsp. lactisNCK20 Lac' R+/M+ spc-4 rcf-5, donor, pTR1041, pTN20 15NCK202 Lac- Hsp- R+/M+, str-15, recipient, pTRK68, previously designated L. lactis subsp. 28

lactis L2FANCK203 Lac- Hsp- R-/M-, str-15, recipient, derivative of NCK202, propagating host for 431 19

and +48NCK204 Lac' Hsp+, str-15, NCK203 transconjugant, pTR2030, pTR1040 19NCK326 Lac' R+/M+, str-15, NCK202 transconjugant, pTN20, pTRK68, pTR1041 This studyNCK340 Lac' R+/M+, spc-4 nif-5, donor, pTRK11 This studyNCK344 Lac' R+/M+, pTRK68, previously designated L. lactis subsp. lactis LMA12 52NCK346 Lac' R+/M+, str-15, NCK203 transconjugant, pTRK11 This studyNCK347 Lac' R+/M+, str-15, NCK203 transconjugant, pTN20, pTRK1041 This studyNCK348 Lac' R+/M+, str-15, NCK202 transconjugant, pTRK11, pTRK68 This studyT-EK1 Lac' Hsp+, donor, pTR1040, pTR2030 35

Bacteriophagesnck203.31 (Q31) Small isometric-headed phage for NCK203 strains 1nck2O3.48 (4)48) Small isometric-headed phage for NCK203 strains, resistant to Hsp+ encoded by 1

pTR2030 (Hsp')Composite HS Combination of 20 individual composite whey samples containing approximately 160 in- M. E. Sanders"

dependently isolated phage; titer of phage that plaque on NCK203 = 106 PFU/mlComposite HR2 Composite HS plus four individual whey samples containing 102 PFU of Hspr phages M. E. Sanders

per mlComposite HR3 Composite HS plus four individual whey samples containing 103 PFU/ml Hspr phages M. E. Sanders

PlasmidspTR1040 Lac' 35pTR1041 Lac' 15pTR2030 Hsp+ R+/M+ Tra+ 35pTRK11 Lac' R+/M' Tra+ 57pTN20C R+/M+ Abi+ Tra+ 15pTRK68 R+/M+ 19

a Smr, streptomycin resistance; Emr, erythromycin resistance; Sp', spectinomycin resistance; Rfr, rifamycin resistance; Nisr, nisin resistance.b Biotechnology Product Division, Miles, Inc.c Neither 448 nor 431 is susceptible to the Abi+ activities encoded by pTN20.

are listed in Table 1. Lactose-fermenting (Lac') Lactococ-cus lactis subsp. lactis strains and their bacteriophages werepropagated in M17-lactose broth at 30°C (60). Lac- strainswere propagated in M17-glucose broth at 30°C. All stockcultures were stored at -20°C in the appropriate mediumwith 10% glycerol. Phages nck2O3.31 (4+31) and nck2O3.48(448) were propagated and their titers were determined on L.lactis subsp. lactis NCK203 (1, 19). Bacteriophage titerswere determined as described previously (60). Several com-posite populations of phages were used in this study. Com-posite HS (HS, Hsp-sensitive phages) was composed of 20individual commercial whey composites and contained ap-proximately 160 independently isolated phages, none ofwhich could form plaques on strain NCK204 containingpTR2030. Phage composite HR2 or HR3 consisted of com-posite HS plus an additional composite containing phagesactive on L. lactis subsp. lactis NCK204 (Hsp+ R+/M+).The phages were designated Hspr and added to the HScomposite (10-6 PFU/ml on NCK203) so that Hspr phages,capable of plaquing on NCK204, would be present at titers of10 PFU/ml (composite HR2) and 103 PFU/ml (compositeHR3).

Conjugation and phage resistance. Conjugation was con-ducted by using agar surface matings, and Lac' transcon-jugants were selected on lactose indicator agar as describedby McKay et al. (45). The following antibiotics were used toselect recipients from mating mixes: spectinomycin, 300pug/ml; rifampin, 50 Ftg/ml; streptomycin, 1,000 ,ug/ml; and

erythromycin, 1.5 ,ug/ml. The phage resistance and sensitiv-ity of transconjugants were evaluated on the basis of effi-ciency of plaquing (EOP) and plaque morphology, as de-scribed previously (35). Plasmid analyses (54) wereconducted on transconjugants to confirm the acquisition ofphage resistance plasmids.SATs. The starter culture activity test (SAT) described by

Heap and Lawrence (12, 14) was conducted in test tubescontaining 10 ml of 11% reconstituted skim milk supple-mented with 1% glucose and 0.25% Casamino Acids. Milktubes were steamed for 1 h, cooled to 30°C, and inoculatedwith 0.2 ml of culture. Each culture was propagated for 16 hat 22°C in milk prior to the inoculation of milk tubes for theactivity test. Starter activity was evaluated on the basis ofthe final pH achieved after the SAT. The time and temper-ature incubation profile of the SAT was 100 min at 30°C, 190min at 40°C, and 100 min at 30°C. Two types of phagesupplements were used in the SATs. In the experimentsusing whey-only supplements, 200 ,l of a phage preparationwas added to freshly inoculated milk cultures. Thereafter,200 ,u of whey from the previous cycle was added at thebeginning of subsequent cycles. In the second procedure,designated phage plus whey supplements, 200 ,ul of theoriginal phage preparation was added initially to the freshlyinoculated milk cultures; at the start of every consecutivecycle thereafter, 100 ,ul of whey from the previous cyclealong with 100 RI of the original phage preparation wasadded. Rotation sequences were terminated when strains

APPL. ENvIRON. MICROBIOL.

on May 15, 2020 by guest

http://aem.asm

.org/D

ownloaded from

ROTATION STRATEGY FOR PHAGE-RESISTANT LACTOCOCCI 367

TABLE 2. Sequence of R/M and Hsp derivatives of L. lactissubsp. lactis NCK203 used in culture rotations

Sequence Cycle no. Systema Strainb Resistance levelc

A 1, 4, 7 R/M I NCK347 10-42, 5, 8 R/M II NCK346 10-33, 6, 9 R/M III NCK344 10-3

B 1, 4, 7 R/M I NCK347 10-42, 5, 8 R/M II NCK346 10-33, 6, 9 H NCK204 <10-9

C 1, 4, 7 H NCK204 <io-92, 5, 8 R/M I NCK347 10-43, 6, 9 R/M II NCK346 10-3

D 1, 4, 7 H NCK204 <10-92, 5, 8 R/M IV NCK326 10-53, 6, 9 R/M V NCK348 10-5

a H, pTR2030; R/M I, pTN20; R/M II, pTRK11; R/M III, pTRK68; R/M IV,pTN20 and pTRK68; R/M V, pTRK11 and pTRK68.

b All strains are derivatives of L. lactis subsp. lactis NCK203.c EOP for small isomeric phage 4)31.

failed to lower the milk pH below 6.0, and high levels ofphage were detected (108 to 109 PFU/ml).

Determination of whey bacteriophage titer. Whey sampleswere prepared by the addition of 0.5 ml of 10% lactic acid tocoagulated cultures after completion of the SAT. The coag-ulated cultures were vortexed, and a 1.5-ml sample wasdispensed into a Microfuge tube. Whey was obtained aftercentrifugation in an Eppendorf Microfuge for 5 min at 10,000rpm. Whey dilutions (10 ,ul of 10-1 to 10-8 dilutions) werespotted on a sensitive indicator lawn ofL. lactis subsp. lactisNCK203 prepared by using an agar overlay technique de-scribed previously (60). Initial phage titers for each cyclewere determined by using the procedure described above formilk with added phage or whey only. Bacteriophage titerswere estimated on the basis of the number of plaques formedper 10-,ul volume of whey preparation and reported to thenearest order of magnitude. The lowest detectable limit was102 PFU/ml.

Starter rotation sequences. The strain rotation sequencesevaluated and order of strains used in each are given in Table2. Phage 431 or 4)48 or phage composites HS, HR2, and HR3were used accordingly with each sequence. Control experi-ments included (i) strains deficient in R/M or Hsp (i.e.,NCK203, no resistance mechanism present), (ii) uninocu-lated milk samples containing phage preparations or wheyonly (initial titer), and (iii) culture only without phage as acontrol for starter activity.

RESULTS

Construction of phage-resistant strains. Strains harboringvarious plasmid-encoded phage defense systems were con-structed for use in starter rotation sequences by usingconjugation or transformation (Table 3). Donor strains car-rying the self-transmissible plasmids pTR2030 (Tra+ Hsp+R+/M+), pTRK11 (Tra+ R+/M+), or pTN20 (Tra+ R+/M+Abi+) were used in conjugation experiments. The Abi sys-tem encoded by pTN20 does not act on phage +31 or +48(8a), so this mechanism did not contribute to the resistanceconferred by NCK203 derivatives bearing pTN20. R/M andAbi/Hsp systems were transferred successfully into thephage-sensitive recipient, L. lactis subsp. lactis NCK203, orits R+/M+ parental strains, namely NCK202 (Lac- R+/M+)and NCK344 (Lac+ R+/M+) (Table 3). Phage +31 did notform plaques on the pTR2030 transconjugant L. lactis subsp.lactis NCK204 (Hsp+ R+/M+; EOP, <10-p). The EOPs of431 on NCK344, NCK346, and NCK347 ranged from 10-'to 10-4. Multiple levels of restriction (EOP, 10-5) weredetected in R+/M+ transconjugants of NCK202 or transfor-mants of NCK344 because of the effect of combiningpTRK68 (R+/M+, naturally residing in NCK202 andNCK344) with pTRK11 (R+/M+) or pTN20 (R+/M+ Abi+).Plasmid pTRK68 (R+/M+), present in L. lactis subsp. lactisNCK344, encoded an R/M system with specificity distinctfrom NCK346 (R+/M+ bearing pTRK11) and NCK347(R+/M+ Abi+ bearing pTN20) (data not shown).

Effect of alternating R/M systems on phage developmentand starter activity. L. lactis subsp. lactis NCK203, thepropagating host for 431, failed in the first cycle whenchallenged with 102 PFU of 431 per ml (data not shown). Ifused individually in repeated cycles, each R+/M+ derivativeof NCK203 (i.e., NCK347, NCK346, or NCK344) failedwithin the second or third cycle because of development ofmodified 431 phages (108 to 109 PFU/ml) (data not shown).R+/M+ strains of L. lactis subsp. lactis NCK347 (bearingpTN20), NCK346 (bearing pTRK11), and NCK344 (bearingpTRK68) were then used in sequence A (Table 2) to deter-mine if alternating R/M mechanisms inhibited phage devel-opment during the SAT. The order of strains used insequence A was established to maximize the degree ofrestriction that the phage would encounter between cycles ofthe different R+/M+-bearing derivatives. Since +31 demon-strated an EOP of 2.4 x 10-4 on L. lactis subsp. lactisNCK347, while NCK346 and NCK344 restricted at thelower levels of 5.7 x 10-3 and 4.2 x 10-3, respectively,NCK347 was chosen to initiate the rotation sequence. Phage431 propagated on L. lactis subsp. lactis NCK347(4)31.NCK347) plaqued on L. lactis subsp. lactis NCK346 ata lower EOP (9.2 x 10-5) than on L. lactis subsp. lactis

TABLE 3. Transfer of phage resistance plasmids into L. lactis subsp. lactis NCK202 (R+/M+) and its R-/M- derivative NCK203 andEOP of phage +31 on the transformants and transconjugants

Donor' Recipient Phage-resistant Relevant plasmid(s)b EOPc Phage resistancerecombinants system

NCK344 pTRK68 (R+/M+) 4.2 x 10-3 R/M IIIT-EK1 NCK203 NCK204 pTR2030 (Hsp+ R+/M+) <10-9 HNCK340 NCK203 NCK346 pTRK11 (R+/M+) 5.7 x 10-3 R/M IINCK20 NCK203 NCK347 pTN20 (R+/M+) 2.4 x 10-4 R/M INCK340 NCK202 NCK348 pTRK11 (R+/M+), pTRK68 (R+/M+) 8.6 x 10-5 R/M VNCK20 NCK202 NCK326 pTN20 (R+/M+), pTRK68 (R+/M+) 3.9 x 10-5 R/M IV

a Donor used in conjugation.b Plasmid(s) that encodes phage resistance in recombinants.c EOP of +31.

VOL. 59, 1993

on May 15, 2020 by guest

http://aem.asm

.org/D

ownloaded from

368 SING AND KLAENHAMMER

CYCLE NUMBER4 5 6

{,- v>_ , RI Rll H RI Rll H RI Rul HRI RII PHAGE RESISTANCE SYSTEM

PHAGE RESISTANCE SYSTEMFIG. 2. Rotation sequence B. The effect of L. lactis subsp. lactis

CYCLE NUMBER NCK204 (Hsp+ R+/M+) on the development of 4)31 during starter1 2 3 4 5 6 7 8 9 culture activity tests is shown. Initial phage 431 titer, 106 PFU/ml

(whey only added after the first cycle). Abbreviations: RI, NCK347;|

INITIALi RII, NCK346; H, NCK204. Final pH values are indicated on the tops

~of the solid bars.

RI RIl Rill RI RIl Rill RI RII RillPHAGE RESISTANCE SYSTEM

FIG. 1. Rotation sequence A. The effect of L. lactis subsp. lactisNCK204 (Hsp+ R+/M+) on phage development during starter cul-ture activity tests is shown. Abbreviations: RI, NCK347; RII,NCK346; H, NCK204. Final pH values are indicated on the tops ofthe solid bars. (A) Initial phage 431 titer, 103 PFU/ml (whey onlyadded after the first cycle; (B) initial phage 4)31 titer, 102 PFU/ml(whey only added after the first cycle).

NCK344 (5.8 x 10-2); therefore, NCK346 was used afterNCK347 in the rotation.

Starter cultures used in sequence A without phage addedproduced acceptable levels of activity (pH 5.0 to 5.3). Thesequence A rotation was then initiated with either 103 or 102PFU of input phage 431 per ml (Fig. 1). Whey only was addedto each subsequent cycle of the SAT. When 103 PFU of 431per ml was present initially, significant levels of phage devel-oped (107 PFU/ml) within the first cycle (Fig. 1A). SequenceA failed after the second cycle, and culture activity wasinhibited (pH 6). When 102 PFU of 431 per ml was presentinitially, some phage development was evident upon the firsttwo cycles. However, phage levels were reduced by the thirdcycle, and culture activity (at pH 5.3 to 5.4) was maintainedthroughout nine cycles (Fig. 1B). These data indicated thatthe rotation of R/M systems encoded by pTN20, pTRK11,and pTRK68 was effective at protecting the background hostL. lactis subsp. lactis NCK203 if the initial levels of phagecontamination did not exceed 102 PFU/ml.

Reduction in phage levels by rotation of strains encodingabortive infection defense systems. In the sequence B rotation(R/M I, R/M II, H), L. lactis subsp. lactis NCK204 (Hsp+R+/M+; bearing pTR2030) was substituted for an R+/M+strain in the third cycle (Table 2). This sequence wasinitiated with 106 PFU of 431 per ml, followed by thewhey-only supplement in subsequent cycles. Minimal cul-ture activity occurred in the first two cycles of the SAT whenthe R+/M+ strains NCK347 and NCK346 were used (Fig. 2).

Phage was not being significantly inhibited, and titersreached high levels of 108 PFU/ml after the first two cycles.When NCK204 (bearing pTR2030) was encountered in thethird cycle, the phage populations were reduced from 106 to103 PFU/ml. Thereafter, rotation of R+/M+ and Hsp+ strainsproceeded through nine cycles of the SAT without loss ofculture activity or buildup of phage.

Since in this experiment L. lactis subsp. lactis NCK204harboring pTR2030 reduced a high initial phage population tolower levels, NCK204 was subsequently used to initiaterotation sequence C (H, R/M I, R/M II). Phage 431 wasadded to the first cycle at levels of 106 PFU/ml; whey onlywas added in each additional cycle. Figure 3 shows thatphage 4)31 populations were reduced from 106 to 103 PFU/mlafter the first cycle. Starter culture activity was maintained(pH 4.9 to 5.4) throughout all rotations, and phage levelswere held to below 102 PFU/ml after the third cycle. Regard-less of whether the R+/M+ strains restricted 4)31 at EOPs of10-3 to 10-4 (Fig. 3A, RI and RII) or 10-5 (Fig. 3B; RIV andRV), acceptable activity occurred through nine cycles of theSAT. If the R+/M+ strains used in sequence C or D(NCK347 and NCK346 or NCK326 and NCK348, respec-tively) were replaced with the 431 propagating host L. lactissubsp. lactis NCK203, culture failure occurred within threecycles as a result of high phage levels (108 PFU/ml) (data notshown).

Inhibition of an Hsp-resistant phage by rotation of Hsp+and R+/M+ strains. Rotation sequence D (H, R/M IV, R/MV) was challenged with phage 448, designated Hspr since itis resistant to the abortive defense mechanism (Hsp) whichis encoded by pTR2030 (1). The pTR2030-encoded R/Msystem still restricts 4)48 at an EOP of 10-3 (1, 19). The othertwo strains in the sequence, L. lactis subsp. lactis NCK326(bearing pTN20 and pTRK68) and NCK348 (bearingpTRK11 and pTRK68), restricted 4)48 at EOPs of 10-'. Therotation sequence D (Table 2) was initiated with 102 and 104PFU of 4)48 per ml, and whey only was added to later cycles.Initial levels of 102 PFU of 4)48 per ml did not disrupt thestarter culture activity of sequence D over nine cycles (Fig.4A). Development of 4)48 over the first three cycles ofsequence D was limited by the sequential activity of thethree R/M systems encountered, including the initial R+/M+activity encoded by pTR2030 in the first cycle. In contrast,initial 4)48 levels of 104 PFU/ml eliminated starter activity bythe sequence D rotation (Fig. 4B). Phage 4)48 developed to

CYCLE NUMBER

A1010

109oiog

108

PFU 107M1 106

1os

104

103

<102

B1010

1og

108

PFU 107YM 106

1o5

104

10 3

C 102

_-| FINAL PFU

5A

I A 5. 5.5 5.4 5.3 54 5.4-

APPL. ENVIRON. MICROBIOL.

on May 15, 2020 by guest

http://aem.asm

.org/D

ownloaded from

ROTATION STRATEGY FOR PHAGE-RESISTANT LACTOCOCCI 369

CYCLE NUMBER1 2 3 4 5 6 7 8 9

A r , .1010 INITIAL PFU

loll FINAL PFU

108_iFINALni

PFU 107

ml106

103 49 ~~~SA<102 5 55.2 A 5.1 5.3 5.4 s.3

H RI Rul H RI Rul H RI RulPHAGE RESISTANCE SYSTEM

CYCLE NUMBER1 2 3 4 5 6 7 8 9

B101B PN T

*INITAL M7

10o_ - FI NAL PFUr108

PFU 107 _

106 _

103<102 4. 54 5.2 5.2 5.3 5.3 5.1 53

H RIV RV H RIV RV H RIV RVPHAGE RESISTANCE SYSTEM

FIG. 3. Rotation sequence C (panel A) and D (panel B). Theeffect of initiating the rotation sequence with Hsp+ L. lactis subsp.lactis NCK204 on phage development during starter culture activitytests is shown. Initial 431 titer, 106 PFU/ml (whey only added afterthe first cycle). Abbreviations: H, NCK204; RI, NCK347; RII,NCK346. Final pH values are indicated on the tops of the solid bars.

high levels (108 PFU/ml) after the first cycle of the sequenceD rotation and caused total culture failure within the fourthcycle. Repeated cycles of solely L. lactis subsp. lactisNCK204 (Hsp+ R+/M+) or NCK203 (4)48-propagating host)after the first NCK204 cycle resulted in culture failure withinthree cycles when initial 448 levels were 102 or 104 PFU/ml(data not shown). Therefore, rotations that cycled R/Mmechanisms could prevent the accumulation of phages re-sistant to the abortive defense mechanism only if low levelsof the Hspr phage were encountered initially.

Starter culture activity in the presence of high-titer indus-trial phage composites. The sequence D rotation (H, R/M III,R/M IV) was evaluated for fermentative activity and longev-ity in the presence of phages isolated from the cheesemakingenvironment (Table 4). High-titer phage composites (com-posites HS, HR2, HR3; Table 1), prepared from commercialwhey samples, were inoculated into the first cycle. In eachsubsequent cycle, whey from the previous cycle was eitheradded alone (-) or added with the original phage composite(+). When sequence D was challenged with composite HS(+), activity was maintained (pH 5.0 to 5.4) over nine cycles(Table 4). If the NCK204 (Hsp+ R+/M+) cycle was followedby a phage-susceptible host (L. lactis subsp. lactisNCK203), failure occurred in the second cycle. Therefore,the sequence D rotation performed adequately in the pres-ence of high-titer industrial phage composites which wereadded to each of the nine cycles of the SAT.The rotation sequence was also challenged with composite

HR2 (Table 4). Again, activity was maintained (pH 5.2 to 5.6)

A1010

109ilo

i08

PFU 10

M 106105

1o4103

<102

CYCLE NUMBER1 2 3 4 5 6 7 8 9

- INITIAL PFU

_ FINAL PFUL

5.0

Ir 4* 5.5 5.3 5.0 5.4 53 5.3 5.4

H RIV RV H RIV RV H RIV RVPHAGE RESISTANCE SYSTEM

CYCLE NUMBER1 2 3 4 5 6 7 8 9

B1010-

log9- 5.7 6.2l 5.0 5.7

PFU

<10104_103-

<102

| INITIAL PUl| FINAL PFU

H RIV RV HPHAGE RESISTANCE SYSTEM

FIG. 4. Effect of +48 (Hspr) on rotation sequences C and Dduring starter culture activity tests. Abbreviations: H, NCK204; RIV,NCK326; RV, NCK348. Final pH values are indicated on the tops ofthe solid bars. (A) Initial titer of +48, 102 PFU/ml (whey only addedafter the first cycle); (B) initial titer of +48, 103 PFU/ml (whey onlyadded after the first cycle).

over nine cycles when whey from each cycle was supple-mented with composite HR2. Rotation sequence D main-tained activity when the phage composite HR3 was not addedto the whey generated after each cycle. Sequence D failedafter three cycles when whey was supplemented with com-posite HR3 (including 103 PFU/ml of Hspr phage) and addedto the beginning of each rotation cycle (Table 4). WhenNCK204 (harboring pTR2030) was followed by cycles of onlythe phage-sensitive host (NCK203) or if NCK204 was usedrepeatedly, starter failure occurred within the second cycle.

DISCUSSION

One of the primary methods used today to limit phagedevelopment in commercial cheesemaking is to rotate starterstrains. Ideally, strains are selected to be sensitive to differ-ent phages (22-24, 31, 42, 43, 61, 62), and, thus, the successof the method depends on the successful identification ofstrains which are not attacked by the same phage strains andspecies. History has proven, however, that it is difficult toidentify truly phage-unrelated strains and accumulate asufficient number of them to run complex rotation programs.Secondly, there are only a limited number of strains whichare available that remain insensitive to phage long enough tomake their introduction into factories worthwhile (12, 21, 39,53). Although some strains are not attacked by existingphages when first introduced, a virulent phage will usuallyappear and build up within the plant (12, 13, 31, 40, 41, 47,49, 61).

VOL. 59, 1993

on May 15, 2020 by guest

http://aem.asm

.org/D

ownloaded from

370 SING AND KLAENHAMMER

TABLE 4. Starter culture activity of strains used in rotation sequence D against whey composites with and without Hsp-resistant phage

Phage b Final pH at cycle no.:compositea Sequence 1 2 3 4 5 6 7 8 9

HS(+) (H, R/M IV, R/M V) 5.0 5.0 5.4 5.0 5.2 5.5 5.0 5.2 5.3HS(+) (H, S, S) 5.0 6.2c

HR2(+) (H, R/M IV, R/M V) 5.5 5.5 5.4 5.2 5.2 5.6 5.2 5.4 5.4HR2(+) (H, S, S) 5.5 6.3cHR2(+) (H, H, H) 5.5 6.4c

HR3(-) (H, R/M IV, R/M V) 5.7 5.5 5.4 5.0 5.2 5.0 5.0 5.2 5.3HR3(+) (H, R/M IV, R/M V) 5.7 5.5 6.4cHR3(-) (H, S, S) 5.7 6.3cHR3(-) (H, H, H) 5.7 6.3c

a Composite HS contains approximately 160 independently isolated phage (titer of phage that plaque on sensitive host NCK203, 106 PFU/ml). Composite HR2consists of composite HS plus 102 PFU/ml of Hspr phage that plaque on NCK204. Composite HR3 consists of composite HS plus 103 PFU/mI of Hspr phage thatplaque on NCK204. (+), whey plus phage composite added to start next rotation; (-), whey only added to start next rotation.

b (H, R/M IV, R/M V), sequence D (Table 2); H, NCK204; S, NCK203 (phage-sensitive host).c Starter failure defined at a final pH of 6.0 to 6.5.

The phage defense rotation strategy (PDRS) evaluated inthis study differs from conventional rotations because a

single bacterial host background is used throughout. There-fore, all of the strains used in the PDRS rotation are

essentially phage related and susceptible to attack by thesame phages and phage species. The PDRS relies on therational incorporation of different phage resistance mecha-nisms into this host background and then rotation of phage-resistant derivatives thereof in a specific sequence whichmaximizes any complementarity between defense mecha-nisms. The model rotation sequence D withstood nine rota-tion cycles when challenged with a commercial whey com-posite containing approximately 160 independently isolatedphages from commercial sources, including Hspr phageswhich are resistant to the abortive mechanism encoded bypTR2030. A rotation sequence based on alternating phageresistance mechanisms within a single isogenic host back-ground is a new concept which can be immediately appliedto commercial starter culture programs.

L. lactis subsp. lactis NCK204 was able to retard phagedevelopment in a manner typical of transconjugants bearingthe pTR2030 plasmid (28, 54, 58). Initiating the sequencewith NCK204 (Hsp+ R+/M+) acted as a cleansing step whicheffectively reduced the number of phage present in the milksystem from an initial level of 106 PFU/ml to a final level of103 PFU/ml after the first cycle. A previous report has shownthat pTR2030 transconjugants derived from L. lactis subsp.cremoris M43, KH, and HP significantly reduced phagelevels from their initial titers during SATs (54). This occurs

since the pTR2030 transconjugants with R+/M+ and Hsp+activities first adsorb phage particles and then either restrictincoming phage DNA or, if that fails, abort the infection atsome later point in the lytic cycle (55). The abortive actionkills the infected cells and in essence entombs any develop-ing progeny phage (34, 36, 56). This phage trap creates a

powerful barrier to phage proliferation and thereby mini-mizes the genetic potential for the appearance of new phagesby mutation, recombination, and host-dependent phage rep-

lication.Genetic addition strategies (6, 11, 26, 34, 48, 52) that

combine abortive and R/M defenses can be used to createstrains containing internal defense systems of unique speci-ficity. The various results generated over the course of thisstudy bear out the importance of both R/M systems, abortivesystems, and their combination when designing derivatives

for the PDRS. Higher levels of phage or the repeated use ofany one R/M system leads to the development of modifiedphage populations capable of halting acid production. Theseobservations confirm previous studies which have demon-strated the limitations of R/M systems in protecting strains inSATs (47, 49). Incorporation of the abortive mechanismHsp+ into one cycle of the rotation complements the activityof R+/M+ by reducing the initial level of phage contamina-tion (106 PFU/ml) to a low level (103 PFU/ml) where R+/M+strains are effective at holding down subsequent phagedevelopment. In turn, R+/M+ strains complement the Hsp+strain by inhibiting the development of Hspr phages. There-fore, the rotation of different mechanisms can serve to limitthe emergence of phages that would otherwise develop uponthe continuous use of any one strain. Rotation sequences canbe rationally designed so that modified phage generated fromone R/M host are restricted at the maximum level in the nextcycle by a host carrying an R/M system of different speci-ficity. Targeting different steps in the phage lytic cycle withdifferent abortive resistance mechanisms could further opti-mize phage inhibition and, therefore, minimize the appear-ance and buildup of new virulent phages within the plant.Another feature of the PDRS is that the same host

background can be used continuously throughout the rota-tion sequence, and this provides a number of importantadvantages. First, highly specialized strains could poten-tially be protected during continuous use in a threateningfermentation environment. For example, this could expandthe usefulness of those industrial strains with optimum levelsof fermentative and organoleptic activities or a novel value-added strain created through biotechnology. Second, pro-longed use of fewer strains via the PDRS will limit thenumber of different lactococcal strains (i.e., host back-grounds) required to maintain cheesemaking. As fewerstrains are used, the diversity of phage populations foundwithin cheese plants drops dramatically (61) and the poten-tial for the emergence of new phages also decreases (12).Third, the appearance of new phages in a plant can bemonitored easily by using the original parental strain as thesensitive indicator for screening whey samples. The devel-opment or buildup of any phages capable of replication onany one of the derivatives of the PDRS could be readilydetected by assays against the original host, which is com-paratively more phage sensitive. Traditional culture rota-tions monitor phage populations in plants by using as many

APPL. ENVIRON. MICROBIOL.

on May 15, 2020 by guest

http://aem.asm

.org/D

ownloaded from

ROTATION STRATEGY FOR PHAGE-RESISTANT LACTOCOCCI 371

as 20 to 50 different indicator strains on a daily basis. The useof a single indicator strain would greatly reduce the amountof work required to monitor emerging phage populations andincrease the sensitivity at which virulent phages against thathost background are detected.The concept of a PDRS developed herein is an innovative

approach for starter rotations. The strategic use of differentmechanism types and specificities enabled us to controlunpredictable phage populations which originated from thecheesemaking environment. This approach to starter protec-tion allowed the use of a single isogenic strain over aprolonged rotation period in a laboratory SAT that mimicsthe conditions encountered in commercial cheesemaking.Delivery of these same phage defense mechanisms into asecond or third host background could provide favorableprotection against phages homologous for that host. Natu-rally occurring plasmids which encode phage defense mech-anisms are abundant in lactococci and can be employedimmediately to construct resistant derivatives for the PDRS.In addition, there are unlimited opportunities to employsome of the more novel mechanisms of phage resistance thatare possible through the application of recombinant DNAtechnologies. Included among these are antisense RNAconstructions (5, 30) and the use of phage origins of replica-tion in trans to compete for essential phage DNA replicationfactors (16). Both of these mechanisms abort phage infec-tions by disrupting the lytic cycle after infection and, there-fore, offer additional opportunities for genetic designs tocapture, trap, and eliminate phages from fermentation envi-ronments. The ability of suspected antisense constructionsto trap and reduce phage populations in SATs was notedrecently by Kim et al. (30). We have noted previously thatabortive mechanisms in general can be used in this capacity(34, 36, 54, 55) and now illustrate how they can serve as onecomponent of a rationally designed rotation strategy toprotect lactococcal strains and prevent the emergence ofnew phages in fermentation environments.

In the future, our ability to utilize desirable and highlyspecialized strains in the industry will be dictated by thephage defense strategies available which can protect theirlong-term and continuous use. The PDRS takes advantage ofour current abilities to construct phage-resistant derivativeswith multiple mechanisms of defense and then use them in arotation scheme designed to prevent the emergence andproliferation of new virulent phages.

ACKNOWLEDGMENTS

This work was supported in part by the Biotechnology ProductsDivision, Miles, Inc., Elkhart, Ind., and the Food IngredientsDivision, Haarmann & Reimer, Inc.We thank Mary Ellen Sanders of Miles, Inc., for providing the

phages and phage composites used in this study and for hersuggestions on the industrial applications for the PDRS. We alsothank C. Hill, D. A. Romero, S. Moineau, and E. Durmaz for theirreview of the manuscript and suggestions during the course of thestudy.

REFERENCES1. Alatossava, T., and T. R. Klaenhammer. 1991. Molecular char-

acterization of three small isometric-headed bacteriophageswhich vary in their sensitivity to the lactococcal phage resis-tance plasmid pTR2030. Appl. Environ. Microbiol. 57:1346-1353.

2. Baumgartner, A., M. Murphy, C. Daly, and G. F. Fitzgerald.1986. Conjugative co-transfer of lactose and bacteriophageresistance plasmids from Streptococcus cremoris UC653.FEMS Microbiol. Lett. 35:233-237.

3. Boussemaer, J. P., P. P. Schrauwen, J. L. Sourrouille, and P.Guy. 1980. Multiple modification/restriction systems in lacticstreptococci and their significance in defining phage-typingsystem. J. Dairy Res. 47:401-409.

4. Chopin, A., M. C. Chopin, A. Moillo-Batt, and P. Langella.1984. Two plasmid-determined restriction and modification sys-tems in Streptococcus lactis. Plasmid 11:260-263.

5. Chung, D. K., S. K. Chung, and C. A. Batt. 1992. AntisenseRNA directed against the major capsid protein of Lactococcuslactis subsp. cremoris bacteriophage 4-1 confers partial resis-tance to the host. Appl. Microbiol. Biotechnol. 37:79-83.

6. Coffey, A., V. Costello, C. Daly, and G. Fitzgerald. 1991.Plasmid-encoded bacteriophage insensitivity in members of thegenus Lactococcus, with special reference to pCI829, p. 131-135. In G. M. Dunny, P. P. Cleary, and L. L. McKay (ed.),Genetics and molecular biology of streptococci, lactococci, andenterococci. American Society for Microbiology, Washington,D.C.

7. Coffey, A. G., G. F. Fitzgerald, and C. Daly. 1991. Cloning andcharacterization of the determinant for abortive infection ofbacteriophage from lactococcal plasmid pCI829. J. Gen. Micro-biol. 137:1355-1362.

8. deVos, W. M., H. M. Underwood, and F. L. Davies. 1984.Plasmid encoded bacteriophage resistance in Streptococcuscremoris SK11. FEMS Microbiol. Lett. 23:175-178.

8a.Durmaz, E., P. L. Higgins, and T. R. Klaenhammer. 1992.Molecular characterization of a second abortive phage resis-tance gene present in Lactococcus lactis subsp. lactis ME2. J.Bacteriol. 174:7463-7469.

9. Froseth, B. R., S. K. Harlander, and L. L. McKay. 1988.Plasmid-mediated reduced phage sensitivity in Streptococcuslactis KR5. J. Dairy Sci. 71:275-284.

10. Gautier, M., and M.-C. Chopin. 1987. Plasmid-determinedsystems for restriction and modification activity and abortiveinfection in Streptococcus cremoris. Appl. Environ. Microbiol.53:923-927.

11. Harrington, A., and C. Hill. 1991. Construction of a bacterio-phage-resistant derivative of Lactococcus lactis subsp. lactis425A by using the conjugal plasmid pNP40. Appl. Environ.Microbiol. 57:3405-3409.

12. Heap, H. A., and R. C. Lawrence. 1976. The selection of starterstrains for cheesemaking. N. Z. J. Dairy Sci. Technol. 11:16-20.

13. Heap, H. A., and R. C. Lawrence. 1977. The contribution ofstarter strains to the level of phage infection in a commercialcheese factor. N. Z. J. Dairy Sci. Technol. 12:213.

14. Heap, H. A., and R. C. Lawrence. 1981. Recent modifications tothe New Zealand activity test for cheddar cheese starters.N. Z. J. Dairy Sci. Technol. 15:91-94.

15. Higgins, D. L., R. B. Sanozky-Dawes, and T. R. Klaenhammer.1988. Restriction and modification activities from Streptococcuslactis ME2 are encoded by a self-transmissible plasmid, pTN20,that forms cointegrates during mobilization of lactose-ferment-ing ability. J. Bacteriol. 170:3435-3442.

16. Hill, C., L. A. Miller, and T. R. Klaenhammer. 1990. Nucleotidesequence and distribution of the pTR2030 resistance determi-nant (hsp) which aborts bacteriophage infection in lactococci.Appl. Environ. Microbiol. 56:2255-2258.

17. Hill, C., L. A. Miller, and T. R. Klaenhammer. 1990. Cloning,expression, and sequence determination of a bacteriophagefragment encoding bacteriophage resistance in Lactococcuslactis. J. Bacteriol. 172:6419-6426.

18. Hill, C., L. A. Miller, and T. R. Klaenhammer. 1991. In vivogenetic exchange of a functional domain from a type II Amethylase between lactococcal plasmid pTR2030 and a virulentbacteriophage. J. Bacteriol. 173:4363-4370.

19. Hill, C., K. Pierce, and T. R. Klaenhammer. 1989. The conju-gative plasmid pTR2030 encodes two bacteriophage defensemechanisms in lactococci, restriction modification (R'/M') andabortive infection (Hsp+). Appl. Environ. Microbiol. 55:2416-2419.

20. Hill, C., D. A. Romero, D. S. McKenney, K. R. Finer, and T. R.Klaenhammer. 1989. Localization, cloning, and expression ofgenetic determinants for bacteriophage resistance (Hsp) from

VOL. 59, 1993

on May 15, 2020 by guest

http://aem.asm

.org/D

ownloaded from

372 SING AND KLAENHAMMER

the conjugative plasmid pTR2030. Appi. Environ. Microbiol.55:1684-1689.

21. Huggins, A. R. 1984. Progress in dairy starter culture technol-ogy. Food Technol. 38:41-50.

22. Huggins, A. R., and W. E. Sandine. 1979. Selection and char-acterization of phage-insensitive lactic streptococci. J. DairySci. 62:70-71.

23. Hull, R. R. 1977. Control of bacteriophage in cheese factories.Aust. J. Dairy Technol. 32:65-66.

24. Jarvis, A. W. 1981. The use of whey-derived phage-resistantstarter strains in New Zealand cheese plants. N. Z. J. Dairy Sci.Technol. 16:25-31.

25. Jarvis, A. W. 1988. Conjugal transfer in lactic streptococci ofplasmid-encoded insensitivity to prolate- and small isometric-headed bacteriophages. Appl. Environ. Microbiol. 54:777-783.

26. Jarvis, A. W. 1989. Bacteriophages of lactic acid bacteria. J.Dairy Sci. 72:3406-3428.

27. Jarvis, A. W., H. A. Heap, and G. K. Y. Limsowtin. 1989.Resistance against industrial bacteriophages conferred on lacto-cocci by plasmid pAJ1106 and related plasmids. Appl. Environ.Microbiol. 55:1537-1543.

28. Jarvis, A. W., and T. R. Klaenhammer. 1986. Bacteriophageresistance conferred on lactic streptococci by the conjugativeplasmid pTR2')30: effects on small isometric-, large isometric-,and prolate-headed phages. Appl. Environ. Microbiol. 51:1272-1277.

29. Josephsen, J. J., and T. R. Klaenhammer. 1990. Stacking ofthree different restriction and modification systems in Lactococ-cus lactis by cotransformation. Plasmid 23:71-75.

30. Kim, S. G., Y.-C. Bor, and C. A. Batt. 1992. Bacteriophageresistance in Lactococcus lactis ssp. lactis using antisenseribonucleic acid. J. Dairy Sci. 75:1761-1767.

31. Klaenhammer, T. R. 1984. Interactions of bacteriophages withlactic streptococci. Adv. Appl. Microbiol. 30:1-29.

32. Klaenhammer, T. R. 1987. Plasmid-directed mechanisms forbacteriophage defense in lactic streptococci. FEMS Microbiol.Rev. 46:313-325.

33. Klaenhammer, T. R. 1989. Genetic characterization of multiplemechanisms of phage defense from a prototype phage-insensi-tive strain, Lactococcus lactis ME2. J. Dairy Sci. 72:3429-3442.

34. Klaenhammer, T. R. 1991. Development of bacteriophage-resistant strains of lactic acid bacteria. Biochem. Soc. Trans.19:675-681.

35. Klaenhammer, T. R., and R. B. Sanozky. 1985. Conjugal trans-fer from Streptococcus lactis ME2 of plasmids encoding phageresistance, nisin resistance, and lactose fermenting ability:evidence for a high-frequency conjugal plasmid responsible forabortive infection of virulent bacteriophage. J. Gen. Microbiol.131:1531-1541.

36. Klaenhammer, T. R., and W. D. Sing. 1991. A novel rotationstrategy using different phage defenses in a single strain starterculture system. J. Dairy Sci. 74:120.

37. Kondo, J. K., and L. L. McKay. 1984. Plasmid transformation ofStreptococcus lactis protoplasts: optimization and use in mo-lecular cloning. Appl. Environ. Microbiol. 48:252-259.

38. Laible, N. J., P. L. Rule, S. K. Harlander, and L. L. McKay.1987. Identification and cloning of plasmid deoxyribonucleicacid coding for abortive phage infection from Streptococcuslactis ssp. diacetylactis KR2. J. Dairy Sci. 70:2211-2219.

39. Lawrence, R. C., H. A. Heap, G. Limsowtin, and A. W. Jarvis.1978. Cheddar cheese starters: current knowledge and practicesof phage characteristics and strain selection. J. Dairy Sci.61:1181-1191.

40. Lawrence, R. C., T. D. Thomas, and B. E. Terzaghi. 1976.Reviews of the progress of dairy science: cheese starters. J.Dairy Res. 43:141-193.

41. Limsowtin, G. K. Y., H. A. Heap, and R. C. Lawrence. 1977. Amultiple starter concept for cheesemaking. N. Z. J. Dairy Sci.Technol. 13:1-8.

42. Limsowtin, G. K. Y., and B. E. Terzaghi. 1976. Phage resistantmutants: their selection and use in cheese factories. N. Z. J.

Dairy Sci. Technol. 11:251-256.43. Marshall, R. J., and N. J. Berridge. 1976. Selection and some

properties of phage-resistant starters for cheesemaking. J. DairyRes. 43:449-458.

44. McKay, L. L., and K. A. Baldwin. 1984. Conjugative 40-megadalton plasmid in Streptococcus lactis subsp. diacetylactisDRC3 is associated with resistance to nisin and bacteriophage.Appl. Environ. Microbiol. 47:68-74.

45. McKay, L. L., K. A. Baldwin, and P. M. Walsh. 1980. Conjugaltransfer of genetic information in group N streptococci. Appl.Environ. Microbiol. 40:84-91.

46. Murphy, M. C., J. L. Steele, C. Daly, and L. L. McKay. 1988.Concomitant conjugal transfer of reduced-bacteriophage sensi-tivity mechanisms with lactose- and sucrose-fermenting abilityin lactic streptococci. Appl. Environ. Microbiol. 54:1951-1956.

47. Pearce, L. E. 1978. The effect of host-controlled modification onthe replication rate of a lactic streptococcal bacteriophage.N. Z. J. Dairy Sci. Technol. 13:166-171.

48. Sanders, M. E. 1988. Phage resistance in lactic acid bacteria.Biochimie 70:411-421.

49. Sanders, M. E., and T. R. Klaenhammer. 1980. Restriction andmodification in group N streptococci: effect of heat on develop-ment of modified lytic bacteriophage. Appl. Environ. Microbiol.40:500-506.

50. Sanders, M. E., and T. R. Klaenhammer. 1981. Evidence forplasmid linkage of restriction and modification in Streptococcuscremoris KH. Appl. Environ. Microbiol. 42:944-950.

51. Sanders, M. E., and T. R. Klaenhammer. 1983. Characterizationof phage-sensitive mutants from a phage-insensitive strain ofStreptococcus lactis: evidence for a plasmid determinant thatprevents phage adsorption. Appl. Environ. Microbiol. 46:1125-1133.

52. Sanders, M. E., P. J. Leonard, W. D. Sing, and T. R. Klaen-hammer. 1986. A conjugal strategy for construction of fastacid-producing bacteriophage-resistant lactic streptococci foruse in dairy fermentations. Appl. Environ. Microbiol. 52:1001-1007.

53. Sandine, W. E. 1989. Use of bacteriophage-resistant mutants oflactococcal starters in cheesemaking. Neth. Milk Dairy J.43:211-219.

54. Sing, W. D., and T. R. Klaenhammer. 1986. Conjugal transfer ofbacteriophage resistance determinants on pTR2030 to Strepto-coccus cremoris strains. Appl. Environ. Microbiol. 51:1264-1271.

55. Sing, W. D., and T. R. Klaenhammer. 1990. Characteristics ofphage abortion conferred in lactococci by the conjugal plasmidpTR2030. J. Gen. Microbiol. 136:1807-1815.

56. Sing, W. D., and T. R. Klaenhammer. 1990. Plasmid-inducedabortive infection in lactococci: a review. J. Dairy Sci. 73:2239-2251.

57. Sing, W. D., and T. R. Klaenhammer. 1991. Characterization ofrestriction and modification plasmids from Lactococcus lactisssp. cremoris and their effects when combined with pTR2030. J.Dairy Sci. 74:1133-1144.

58. Steenson, L. R., and T. R. Klaenhammer. 1985. Streptococcuscremoris M12R transconjugants carrying the conjugal plasmidpTR2030 are insensitive to attack by lytic bacteriophages. Appl.Environ. Microbiol. 50:851-858.

59. Steenson, L. R., and T. R. Klaenhammer. 1986. Plasmid heter-ogeneity in Streptococcus cremoris M12R: effects on proteo-lytic activity and host-dependent phage replication. J. Dairy Sci.69:2227-2236.

60. Terzaghi, E. A., and W. E. Sandine. 1975. Improved medium forlactic streptococci and their bacteriophages. Appl. Microbiol.29:807-813.

61. Thunell, R. K., W. E. Sandine, and F. W. Bodyfelt. 1981.Phage-insensitive, multiple-strain starter approach to cheddarcheesemaking. J. Dairy Sci. 64:2270-2277.

62. Whitehead, H. R., and G. J. E. Hunter. 1947. Bacteriophage incheese manufacture. Contamination from farm equipment. J.Dairy Res. 15:112-120.

APPL. ENVIRON. MICROBIOL.

on May 15, 2020 by guest

http://aem.asm

.org/D

ownloaded from