MONITORING BACTERIOLOGICAL QUALITY OF PRECOOKED LASAGNA

Abstract

The effects of two preparation methods, two packaging materials, and modified atmosphere condition, on the bacteriological quality and safety of precooked lasagna, were studied. Microbiological analyses focused on Salmonella spp., Listeria monocytogenes, Bacillus cereus, Escherichia coli, Staphylococcus aureus, spores of sulphite-reducing clostridia, and on mesophilic and psychrotrophic aerobic bacteria, total coliforms and lactic acid bacteria. In total, 81 samples were examined, for up to 28 days. No pathogenic microorganisms were detected. After 28 days, spoilage-hygiene markers levels were non-compliant to reference standards. In conclusion, lasagna prepared with hot method in polyethylene packages were acceptable for up to 21 days. [PUBLICATION ABSTRACT]

Headnote

ABSTRACT

The effects of two preparation methods, two packaging materials, and modified atmosphere condition, on the bacteriological quality and safety of precooked lasagna, were studied. Microbiological analyses focused on Salmonella spp., Listeria monocytogenes, Bacillus cereus, Escherichia coli, Staphylococcus aureus, spores of sulphite-reducing clostridia, and on mesophilic and psychrotrophic aerobic bacteria, total coliforms and lactic acid bacteria. In total, 81 samples were examined, for up to 28 days. No pathogenic microorganisms were detected. After 28 days, spoilage-hygiene markers levels were non-compliant to reference standards. In conclusion, lasagna prepared with hot method in polyethylene packages were acceptable for up to 21 days.

- Keywords: precooking, modified atmosphere packaging, lasagne, catering -

1. INTRODUCTION

In Italy it is estimated that state school catering services serve more than 2,700,000 students daily. This represents a huge "social restaurant" in which the State, through local authorities, basically chooses the organization, raw materials, and the menus (Emilia-Romagna, Italy, 2009; Conferenza Permanente Stato-Regioni, 2010). Considering the complexity of the organizations involved, the vulnerability of the people who receive the foodservices, and the consequences of possible adverse events, it is clear that catering services are of great strategic importance in the program to develop the right nutrition and to ensure food safety and nutritional quality.

In Italy the traditional "cook-serve" foodservice system is the most common in school catering (CIAPPELLANO, 2009), in which ingrethents are assembled and food is produced on-site, kept either heated or chilled for up to two hours, and served to consumers.

Nevertheless, in the school kitchens that prepare thousands of meals per day there are some dishes that cannot be prepared using the cookserve method, due to the complexity of the recipe and the long period of time requested for assembly and cooking. In these cases the caterers often purchase already precooked, chilled or frozen dishes from outside suppliers, such as lasagna.

Lasagna is a classic Italian pasta casserole dish, which consists of alternate layers of pasta, cheese and often other ingrethents such as minced beef with tomato sauce. It is one of the most well-liked recipes in Italian school foodservices (DRAGONI etat, 2005).

Modified atmosphere packaging (MAP) in combination with chill storage (<4°C) is a well-known food preservation method, but in the Italian school catering is still not widespread in meals preparation systems (CIAPPELLANO, 2009). The right combination of gases (CO2, N2 and O2) in the headspace of food packs suppresses the microbial flora of perishable foods that normally develop under aerobic conditions (PATSIAS etol, 2003). A minimum CO2 concentration of 20-30% is necessary to produce inhibitory effects (STILES, 1991). Chilling slows the deterioration of stored foods but if the atmosphere surrounding the product is also modified to reduce oxygen concentration, the shelf-life is increased considerably because of a further reduction in the rate of chemical oxidation by oxygen and in the growth of aerobic microorganisms (PHILLIPS, 1996). Nevertheless, concerns have been expressed that the increase in shelf-life of MAP products provides sufficient time for human pathogens to multiply to levels which render the food unsafe while still edible (JAY, 1992).

Pathogens that are able to multiply at chill temperatures (-1° to +70C) (RAY, 2004) and that are able to growth in very low levels of oxygen as psychrotrophic Clostridium, botulinum and Listeria monocytogenes (minimum growth temperature: 0° to 1°C) (RAY, 2004) are of a particular concern (PHILLIPS, 1996; RAY, 2004). Some of the major microbiological hazards associated with modified atmosphere packaging are: 1) Clostridium botulinum type E, which is capable of growth and toxin production at 3°C; 2) Listeria monocytogenes, enterotoxigenic Escherichia coli and spore-formers such as Bacillus cereus can survive inadequate heat and then grow during the chilled storage of the product; 3) any leaks in the seal and packaging material can lead to post- thermal processing contamination by pathogens (????, 2000). However, the use of MAP with precooked meals that are subsequently cooked is considered less hazardous because cooking (if correctly carried out) kills vegetative pathogens such as Listeria monocytogenes, Salmonella sop., Escherichia coli. Staphylococcus aureus, Bacillus cereus, and Clostridia (HOTCHKISS, 1988).

The aim of this study was to evaluate and compare the bacteriological quality and safety of refrigerated precooked lasagna, experimentally prepared by a traditional cook-serve centralized school kitchen in Lombardy, Italy. This study was carried out testing the efficacy of two production processes and two types of packaging materials, under a modified atmosphere, and stored at a refrigerator temperature (cold room - Angelo Po Grandi Cucine S.p.A., Italy - set at 2°C) for up to 28 days. The ultimate goal was to determine the most suitable production method and type of packaging to ensure as high a level of microbial safety as possible. The microbiological analysis focused on pathogenic and potential-pathogenic microorganism markers [Salmonella spp., Listeña monocytogenes, Bacillus cereus, Escherichia coli, Staphylococcus aureus, and spores of sulphite-reducing Clostridia); as well as spoilage-microorganisms and hygienic markers: mesophilic aerobic and psychrotrophic bacteria, total coliforms and lactic acid bacteria.

2. MATERIALS AND METHODS

This study was conducted from June 2010 to July 2010, involving a cook- serve kitchen of an Italian school catering establishment in Lombardy, which provides 3,000 meals a day for 3 to 14 year-old children.

2.1 Preparation of lasagne

The lasagna samples were prepared on 23 June 2010. The production was conducted in a separate area of the establishment. The ingredients consisted of: a) Bolognese sauce (40%) (Table 1), b) Béchamel sauce (40%) (Table 2) and e) lasagna sheets (20%) (durum wheat, wheat flour type 00, pasteurized whole egg - 20% - and water). Bolognese and Béchamel sauces were cooked in the on-site, pasta sheets were provided precooked and frozen (Food Valley S.r.L, Florence, Italy). The lasagna consisted of five layers of pasta sheets interleaved with three layers of Bolognese sauce and three layers of Béchamel.

The two lasagna preparation methods and the two types of packaging materials are shown and described in Fig. 1.

2.2 Storage conditions and sampling

The lasagna (n = 81) was stored in a cold room set at 2°C for 28 days. The temperature of the cold room was monitored every 30 minutes by a temperature recorder (iLOG, Escort Data Loggers Inc., Usa). For the purpose of this study, precooked lasagna portions were divided into four groups according to their preparation and packaging methods. Group A (n = 15): HSL method and CPET/APET packaging; Group B (n = 15): CSL method and CPET/APET packaging; Group C (n = 15): HSL method and CELLPET packaging; Group D (p= 15): CSL method and CELL-PET packaging. During the storage, the analysis of the headspace gas concentration and sampling for the microbiological tests were performed on day 0 and after 7, 14, 21 and 28 days of storage. At time 0, the frozen precooked pasta sheets, and the Béchamel and Bolognese (? = 9) sauces after cooking were packaged and analysed separately. At time 28, three samples per group (n= 12) were analysed after cooking at 1800C for 30 minutes. The cooking was carried out using a convection oven (Zanussi S.p.A., Italy) set at "dry cooking". All samples were transported to the laboratory in containers with ice and were analysed the same day.

2.3 Headspace gas analysis

Gas concentrations in the headspace of packages (CO2 and O2) were determined using an OXYBABY®-M headspace analyzer (WITT GasetechniK GmbH & CO., England).

2.4 Microbiological analysis

An analytical unit (10 g) was aseptically taken from each sample unit, added to 90 mL of sterile diluent solution (0.85% NaCl and 0.1% peptone), and homogenized in a stomacher 400 (Colworth, UK) for 1 min at room temperature and then serial 10-fold dilutions were prepared in a sterile saline.

Psychrotrophic aerobic plate counts (PAPC) were determined using a pour plate technique on Plate Count Agar (Oxoid, Basingstoke, UK); plates were incubated at 100C for 10 days. This medium was chosen since it is a well-known and frequently applied tool for the enumeration of bacteria in food quality control programs in the food industry (ESPE et al, 2004; BERNARDI et al, 2009), and it is the medium indicated by official methods (ISO 17410: 2001). Mesophilic aerobic plate counts (MAPC) were enumerated using a Petrifilm(TM) Aerobic Count (3M(TM), SL Paul, Minnesota, USA), following the AFAQ /AFNOR 3M 0 1 / 1 -09/89 method. Petrifilm(TM) plates were also used to determine total coliforms (TC), Escherichia coli (EC), Staphylococcus aureus (SA), using the following methods AFNOR 3M 0 1 /2-09/89A AFAQ/AFNOR 3M 01/08-06/01 and AFNOR 3M 01/9-04/03 respectively. Lactic acid bacteria (LAB) were enumerated on de Man-Rogosa-Sharpe agar (Oxoid, Basingstoke, UK), atpH 5.5. Plates were incubated at 300C for 48 h under anaerobic conditions in an anaerobic jar (Oxoid, Basingstoke, UK) with an AnaeroGen(TM) sachet oxygen absorber (Oxoid, Basingstoke, UK); Salmonella spp. detection was in accordance with ISO 6579:2002 Cor. 1:2004. The detection of Listeria monocytogenes were performed in accordance with AFNOR BRD 07/409/98 and AOAC No. 060402 2006. For Bacillus cereus enumeration, 0. 1 mL of each dilution sample was put onto polymixin-piruvate-egg yolkmannitol-bromothymol blue agar (PEMBA Oxoid, Basingstoke, UK) with 50,000 IU of polymixin per litre and egg yolk emulsion (Oxoid, Basingstoke, UK). The plates were air dried and incubated at 300C for 24 to 48 h (TESSI et al, 2002). For the isolation and enumeration of spores of sulphitereducing Clostridia (SSC), the first step was the heat treatment of the tubes containing the first decimal dilution in a thermostatically controlled water bath at 80°±0. 1°C for exactly 10 min after the temperature reached 800C in a control tube, in order to eliminate vegetative cells. In the second step, spores were enumerated by the pour plate method (SWANSON et al, 1992) onto tryptose-sulfite-cycloserine (TSC) agar without egg yolk (CM587, Oxoid, Basingstoke, UK). The agar was supplemented with a 1% filter-sterilized Dcycloserine (Oxoid, Basingstoke, UK) solution (4% wt/vol.). An amount of 0. 1 mL of heated dilution was put onto empty sterile Petri dishes, poured plates were overlaid with a thin layer (5 mL) of freshly prepared TSC agar (supplemented with D-cycloserine as previously described). Upon solidification of the TSC overlay, plates were placed in anaerobic jars (Oxoid, Basingstoke, UK) with an AnaeroGen(TM) sachet oxygen absorber (Oxoid, Basingstoke, UK) and incubated at 37.00C for 48 h. Characteristic black colonies, if present, were enumerated. All analyses were performed in triplicate.

2.5 Evaluation of microbial safety

Codex Alimentarius Commission guidelines (1997) were followed for evaluating the microbial safety of samples: L. monocytogenes (not detected in 25 g), Salmonella spp. (not detected in 25 g), E. coli (satisfactory <1.00 log10 CFU/g, unsatisfactory >1.00 log10 CFU/g), B. cereus (satisfactory <2.00 log10 CFU/g, unsatisfactory >2.00 log10 CFU/g), S. aureus (satisfactory <2.00 log10 CFU/g, unsatisfactory >2.00 log10 CFU/g), spores of sulphite-reducing Clostridia (satisfactory < 1.00 log10 CFU/g, unsatisfactory >1.00 log10 CFU/g). Mesophiles and psychrotrophic aerobic plate counts, total coliforms and lactic acid bacteria provide a general estimation of the total number of microorganisms on the product, and are very helpful in estimating/comparing the microbial quality of each group of samples analysed (KOKKINAKIS et al, 2007). In our study we established the following acceptance limits: MAPC/PAPC = 6.00 log10 CFU/g; TC = 3.00 log10 CFU/g and LAB = 6.00 log10 CFU/g.

3. RESULTS AND CONCLUSIONS

Salmonella spp. and Listena monocytogenes were not detected in any of the samples analysed. Escherichia coli was <1.00 log10 CFU/g, Bacillus cereus was <2.00 log10 CFU/g, spores of sulphite-reducing Clostridia were <1.00 1Og1 0 CFU/g in all the samples examined. In Groups A and D LAB counts were lower than the acceptance limit up to 28 days of storage (mean counts ranged from <2.00 at day zero to 5.66±0.02 log10 CFU/g at day 28), while in Group B and C LAB counts exceeded the acceptance limit on 28th day of storage (respectively with 6.98±0.02 and 6. 11±0.07 log10 CFU/g).

Staphylococcus aureus was found with a frequency of contamination of 37% (mean counts varied from < 1.00 to 2.88±0.08 log10 CFU/g). In half of the positive samples S. aureus counts were unsatisfactory (mean values were > 2.00 log10 CFU/g), and the unsatisfactory counts were found in Groups B and D (cold sauce layering) from day 14 (Group B) and from day 28 (Group D), respectively. No difference in contamination levels was found between CELL-PET and CPET/ APET packages.

It is well known that S. aureus lives on humans, food equipment, environmental surfaces, and animals, and exists in the nasal passages, throats, hair, and skin of 50% or more of healthy individuals. This bacterium is highly vulnerable to heat treatment and nearly all sanitizing agents (YOON et al., 2008). For these reasons we can assume that the presence of S. aureus may be an indication of an environmental contamination during the filling process. In case of the hot fill system S. aureus has been eliminated by the high temperatures of the sauce, in the cold fill system S. aureus has not been eliminated.

The evolution of microbial populations (MAPC, PAPC, TC, LAB) in the four Groups are shown in Figs. 2-5. The trend in the percentage of carbon dioxide measured in the headspace of packages is shown in Fig. 6. During the period of 28 days, the mean temperature in the cold room recorded by the data logger was 2. 1°C.

As regards the enumeration of mesophilic (MAPC) and psychrotrophic (PAPC) microorganisms, we found that the counts were the same order of logarithmic magnitude (Figs. 2 and 3). We can thus hypothesize that almost all the mesophilic bacteria detected had the capacity to grow at refrigeration temperatures.

At day 0, precooking seemed the most important factor to affect initial bacterial populations, as mean mesophilic and psychrotrophic counts ranged from 2.00±0.04 to 3.98±0.02 log10 CFU/g and total coliforms counts (Fig. 4) ranged from <1.00 to 2.87±0.03 log10 CFU/g.

The results of the tests performed separately on the three elements of lasagna (at day 0), carried out immediately after packaging, snowed that mean MAPC and PAPC in frozen precooked sheets were respectively 4.00±0.03 log10 CFU/g and 4.04±0.06 log10 CFU/g, in Béchamel sauce MAPC and PAPC were respectively 2.60±0.06 log10 CFU/g and 2.00+0.05 log10 CFU/g. In the Bolognese sauce they were <1.00 log10 CFU/g, and all the other microbiological parameters (TC, SA, LAB) were lower than the detection limits.

At day 14, the inhibitory effect of CO2 was particularly significant as the populations were substantially unvaried in Groups A, B and C; only in Group D (CELL- PET packaging and CSL method) mean values of MAPC, PAPC and total coliforms counts increased by 1.25-1.57 log10 units, and TC mean counts were unsatisfactory according to the acceptance limits (mean value was 3.63±0.06 log10 CFU/g).

At day 21, MAPC, PAPC and TC mean counts exceeded the acceptance limits only in Group D.

At day 28, the populations detected were higher than the acceptance limits in all the four Groups analysed (mean MAPC and PAPC ranged from 6.18±0.02 to 8.08±0.07 log10 CFU/g, TC mean values varied from 3. 1 1±0.03 log10 CFU/g in Group C, to 4.95±0.03 log10 CFU/g in Group D).

The results we obtained confirm that the CO2 concentration inhibited the growth of microorganisms as long as a sufficient concentration of dissolved CO2 was maintained on the surface of the lasagna. As shown in Figure 6, the carbon dioxide percentage decreased rapidly until day 7, then decreased more slowly up to day 28. C and D were the groups in which the CO2 decreased to the lowest levels.

These data indicate that probably the heat sealing of CELL-PET containers with a PET top film could be less effective over time compared to the polyethylene containers. The inefficiency of heat-sealing was also demonstrated by the percentage of oxygen (Fig. 7) measured in the headspace. In fact, during the 28 days of storage, in Group C oxygen increased from 2.3 to 18.8%, in Group D it increased from 0.0 to 2.9%, while in Groups A and B it increased respectively from 0.2 to 2.2% and from 0.0 to 1.3%. The level of oxygen may have inhibited the surface growth of pathogenic anaerobic bacteria, which were not detected in this investigation. No difference in aerobic bacterial counts was found between higher and lower percentage oxygen packages, most probably because the bacterial populations were also able to grow also in an environment with an oxygen deficiency (<1%).

At day 28, the results on the samples analysed after cooking showed almost a complete destruction of microbial populations, in all the groups, thus demonstrating the efficacy of cooking timing and temperature.

In conclusion, this study investigated the possible extension of the shelf-life of refrigerated precooked lasagna prepared by using two different processes and packaged using two materials, under MAP. The microbiological results suggest that the potential risk of the growth of pathogenic microorganisms is minimal.

Both packaging methods (CPET-APET and CELL-PET) ensured a sufficiently long shelf-life for most catering purposes. The only difference worth noting was that cellulose had higher values of oxygen in the headspace, which could affect its future use in catering due to the greater probability of aerobic bacterial growth. In conclusion, the use of MAP, and both the HSL/CSL methods associated with CPET-APET packaging and the HSL method associated with CELL- PET containers appeared to be an optimal solution for increase the shelf-life up to 2 1 days.

The results obtained could be useful to implement new production processes within the widespread Italian cook-serve school foodservice establishments, which currently, due to the large number of consumers to be served, acquire ready-to-cook frozen lasagna from outside suppliers in order to optimize time and labor.

There are several direct and indirect advantages from the implementation of this new production system in the largest cook-serve catering establishments: a) greater flexibility in menus; b) possibility of customize the recipes for ethnic and religious minorities (for example using poultry meat instead of beef), and for children suffering from food allergies or intolerances; c) meet the contemporary trend for consumption of fresh rather than frozen products.

ACKNOWLEDGEMENTS

The authors would like to thank Elena Fucile, Food Technologist, and Loris Vaccari, Production Manager, for their precious technical assistance.

References

REFERENCES

AOAC International The Scientific Association Dedicated to Analytical Excellence®. 2010. Validation Certificate No. 060402. Microgen Listeria-ID Kit for confirming the presence of Listeria monocytogenes.

Association Française De Normalisation (AFNOR) 20 1 0. Validation Certificate N. 3M-01/2-09/89A. 3M(TM) Petrifilm(TM) Conform Count Plate. Application for the enumeration of conform colonies with and without gas (as compared to the ISO 4832 VRBL Method).

Association Française De Normalisation (AFNOR) 2010. Validation Certificate N. BRD 07/4-09/98. Application for detection of Listeria monocytogenes (as compared to the NF EN ISO 1 1290-1 including the amendment Al (2004) Method).

Association Française De Normalisation (AFNOR) 20 1 0. Validation Certificate N. 3M-0 1/9-04/03. 3M(TM) Petrifilm(TM) Staph Express Count System. Application for the enumeration of coagulase positive Staphylococcias compared to the EN ISO 6888-1 Method).

Association Française pour l'Assurance de la Qualité (AFAQ) et Association Française De Normalisation (AFNOR) 20 1 0. Validation Certificate N. 3M-0 1/1-09/89. 3M(TM) Petrifilm(TM) Aerobic Count Plates. Application for the Enumeration of Total Viable Count (as compared to the ISO 4833 Method).

Association Française pour l'Assurance de la Qualité (AFAQ) et Association Française De Normalisation (AFNOR) 2010. Validation Certificate N. 3M-0 1/08-06/01. 3M(TM) Petrifilm(TM) Select E. coli Count Plate. Application for the enumeration of Escherichia coli (as compared to the ISO 16649-2 E. coli Method).

Bernardi C, Ripamonti B., Campagnoli ?., Stella S. and Cattaneo P 2009. Shelf-life of vacuum packed Alaskan, Scottish and Norwegian cold-smoked salmon available on the Italian market. Int. J. Food Sci. Technol., 44: 2538-2546.

Ciappellano S. 2009. In: "Manuale della Ristorazione". Casa Editrice Ambrosiana, Milan, Italy.

Codex Alimentarius Commission 1997. Joint FAO/WHO Food Standards Programme, Codes Committee on Food Hygiene. In: "Food Hygiene", Supplement to Volume JR1997. Principles for the establishment and application of microbiological criteria for foods. CAC/GL 21-1997. Secretariat of the Joint FAO/WHO Food Standards Programme. Rome: Food and Agriculture Organization of the United Nations.

Conferenza Permanente Stato-Regioni 2010. In: "Linee di indirizzo nazionale per la ristorazione scolastica". Rome, Italy, 29 April 2010. Available from: http://www. statoregioni.it/Documenti/DOC_026408_2CU.pdf.

Direzione Generale Sanità e Politiche sociali, Regione Emilia-Romagna 2009. In: "Linee strategiche per la ristorazione scolastica in Emilia-Romagna". Bologna, Italy. Available from: http://www.saluter.it/wcm/saluter/pubblicazioni/tutte_le_pubblicazioni/contributi/contributi.htm.

Dragoni LG., Vallone L. and Balzaretti CM. 2005. Alunni in cattedra: indagine sulle metodologie di rilevazione della Customer Satisfaction nella ristorazione scolastica. Ristorando, 5: 54-56.

Espe M., Kiessling A., Lunestad B.T., Torrissen O.J. and Bencze Rora A.M. 2004. Quality of cold-smoked salmon collected in one French hypermarket during a period of 1 year. LWT- Food Sci. Technol., 37: 617-638.

Hotchkiss J.H. 1988. Experimental approaches to determining the safety of food packaged in modified atmospheres. Food Technol., 42: 55-64.

International Organization for Standardization (ISO 17410:2001) 2001. Microbiology of food and animal feeding stuffs - Horizontal method for the enumeration of psychrotrophic microorganisms. Geneva: International Organization for Standardization.

International Organization for Standardization (ISO 6579:2002 Cor. 1:2004) 2004. Microbiology of food and animal feeding stuffs. Horizontal method for the detection of Salmonella spp. Geneva: International Organization for Standardization.

Jay J.M. 1992. Microbiological Food Safety. Crit. Rev. Food Sci., 31: 177-190.

Kokkinakis E., Boskou G., Fragkiadakis G.A., Kokkinaki A. and Lapidakis N. 2007. Microbiological quality of tomatoes and peppers produced under the good agricultural practices protocol AGRO 2-1 & 2-2 in Crete, Greece. Food Control, 18: 1538-1546.

Nyati H. 2000. An evaluation of the effect of storage and processing temperatures on the microbiological status of sous vide extended shelf-life products. Food Control, 11:471-476.

Patsias A., Chouliara L, Badeka A, Sawaidis LN. and Kontominas M.G. 2003. Shelf-life of a chilled precooked chicken product stored in air and under modified atmospheres: microbiological, chemical, sensory attributes. Food Microbiol., 23: 423-429.

Phillips CA. 1996. Review: Modified Atmosphere Packaging and its effects on the microbiological quality and safety of produce. Int. J. Food Sci. Technol., 31: 463-479.

Ray B. 2004. Fundamental Food Microbiology 3rd Edition. CRC Press LLC, Florida, United States of America.

Stiles M.E. 1991. Scientific principles of controlled /modified atmosphere packaging. In: Ooraikul B. and Stiles M.E. (Eds.), "Modified Atmosphere Packaging of Food". Ellis Horwood Ltd., England, pp. 18-15 and 118-147.

Swanson K.M.J. , Busta F.F., Peterson E.H. and Johnson M.G. 1992. Colony count methods. In: Vanderzant C, Splittstoesser D.F (Eds.) "Compendium of methods for the microbial examination of foods". American Public Health Association, Washington D. C, 75-95.

Tessi M.A., Aringoli E.E., Pirovani M.E., Vincenzini A.Z., Sabbag N.G., Costa S.C., Garcia CC, Zannier M.S., Silva E.R. and Moguilevsky M.A. 2002. Microbiological quality and safety of ready-to-eat cooked foods from a centralized school kitchen in Argentina. J. Food Prot., 65(4): 636-642.

Paper received July 13, 2011 Accepted November 4, 2011