Page 1 of 6

APIACATA 38 (2004)352-357 352

STORAGE PROTEINS IN WINTER HONEY BEES

G. W. Otis1

; D. E. Wheeler2

; N. Buck2

; H. R. Mattila1

1

University of Guelph, Guelph, Ontario, Canada N1G 2W1 2

University of Arizona, Tucson, Arizona, USA 85718

ABSTRACT

Honey bee colonies in temperate climates begin brood rearing in late winter before

floral resources are available. Protein required for brood rearing comes from pollen

stored in combs as well as proteins stored internally in the bees' bodies. We studied

the changes in amounts of several proteins that could serve as internal storage

compounds in fall-emerging bees that contributed to the wintering population.

Vitellogenin, a known storage protein, increased from none in newly emerged bees to

~60 (range: 10-200)ug/abdomen in winter bees (60-day-old bees sampled in late

November). Probable jelly proteins from bee heads were present in minor amounts.

A third previously unreported protein from adult honey bee abdomens was a six-unit

storage protein, probably arylphorin. Like vitellogenin, it was nonexistant in newly

emerged bees and reached an average of 75 ug/bee by late November.

Although the two hives from which we sampled bees were randomly selected, there

were highly significant colony-related differences in amounts of protein in the bees.

Bees from one colony consistently had higher levels of all proteins studied.

We have demonstrated the existence of a previously unknown storage protein in

wintering bees that is probably arlyphorin. Although wintering bees have significantly

elevated amounts of vitellogenin and arylphorin, the amounts of protein present in the

bodies of wintering bees are not sufficient to rear large amounts of brood. We

suggest that their primary function may be to allow colonies to continue brood rearing

when extremely cold periods prevent bees from leaving the cluster to feed on pollen.

Other possible functions are suggested.

Keywords: storage proteins, vitellogenin, winter bees, arylphorin

INTRODUCTION

A key adaptation that enabled tropical Apis to colonize temperate regions was the

ability to survive cold winters. Several traits must co-occur to make this possible:

regulation of the cessation of brood rearing in fall and its initiation in the late winter

prior to the availability of floral food resources; the ability of individual worker bees to

live for many months; storage of large amounts of honey for thermoregulation; and

storage of sufficient protein for late winter and early spring brood-rearing (Seeley

1985).

All protein in a colony of bees is ultimately derived from floral pollen. Some pollen is

fed directly to older larvae. Additional pollen is ingested by adult worker bees,

converted to jelly, and fed to larvae of all castes as well as to older adult workers,

Page 2 of 6

APIACATA 38 (2004)352-357 353

drones, and queens (Crailsheim 1992). Protein from pollen is required for the

development of glands (e.g., hypopharyngeal glands, wax glands) during the

behavioural ontogeny of worker bees. Finally, some protein is stored internally within

worker bees, primarily in the fat body, haemolymph, and hypopharyngeal glands

(Amdan and Omholt 2002).

In late summer and fall in temperate regions, bees store pollen along with honey.

They also consume pollen and develop hypertrophied hypopharyngeal glands and

large fat bodies containing globules of protein, traits related to the development of

long-lived winter bees (Ribbands, 1953). Vitellogenin, a known internal storage

protein, increases in amount in wintering bees along with several other proteins

(Amdan and Omholt 2002). When brood rearing commences in mid-winter, bees

must obtain protein for larvae from stored pollen and internal protein reserves.

A wide variety of insects synthesize special storage proteins during periods of

resource abundance. These are retained until the time and setting required for egg

production or other functions including cuticle formation, transport of organic

compounds, and humoral immune defense (Burmester 1999). For example, queen

ants of several species have large amounts of large hexameric (e.g., comprised of

six subunits) proteins that are digested to produce food for their first larvae during

claustral colony founding (Wheeler and Martinez 1994, 1995, Wheeler and Buck

1995). Adult worker honey bees have been documented as containing Hex70a in

relatively large amounts (Danty et al. 1998), but details about the amounts and its

function are lacking. In addition to hexamerins, there are additional proteins that may

serve as protein storage molecules, including vitellogenin which is a very high density

lipoproteins (VHDLs) (Wheeler and Buck 1995, Amdan and Omholt 2002). There has

been little research into internal sources of protein in honey bees with the recent

exception of Amdan and Omholt (2002). Our study investigated the amounts of

stored protein in developing winter bees, and discusses their likely usage within the

colony.

MATERIALS AND METHODS

For this study two colonies (#77 and #91) were selected at random in late summer at

the University of Guelph, Ontario. Frames of emerging bees were removed,

incubated overnight, and large cohorts of bees <1 day old were paint marked and

returned to their colonies of origin on 4 September and 30 September, 2000.

Previous studies indicated that these would become “summer” and “winter” bees,

respectively. Samples of ten bees from each cohort were collected at various ages

(e.g., 0, 2, 5, 7, 11, 14 days of age) and frozen at –80 C. On November 30, when all

bees in the colonies were >60 days of age, an additional sample of “winter bees” was

obtained. The bees were shipped on dry ice to the University of Arizona where the

methods previously reported by Wheeler and Martinez (1994) were used to analyze

proteins present in the bees. In summary, the method involved removal of the gut

and sting, after which the head, thorax and abdomen were homogenized individually

in tris-buffered saline in the presence of several protease inhibitors. After

centrifugation, the supernatent was analyzed with SDS-PAGE (polyacrylamide gel

electrophoresis). After electrophoresis, a numerous bands were evident on the gels.

By their size and location in the bee, several of these were identified and quantified:

lipophorin, vitellogenin, a putative storage protein (probably arylphorin), an abundant

thorax protein (probably muscle-related), and a putative jelly protein.

Page 3 of 6

APIACATA 38 (2004)352-357 354

RESULTS

LIPOPHORIN

Lipophorin is a lipid-carrying protein that is generally a good indicator of the physical

condition of an insect/colony. There was a significant colony effect (p=0.02),

indicating that bees from Colony 77 had significantly more lipophorin than bees from

Colony 91. There was no effect of age (p=0.59) or age*colony effect (P=0.86).

Although we selected colonies at random within the home apiary, this analysis and all

others (below) indicate that Colony 77 was nutritionally better off.

VITELLOGENIN

Vitellogenin (VG) is a VHDL that acts as a storage protein that appears to be

involved in various metabolic functions including the production of hypopharyngeal

gland secretions (Amdan and Omholt, 2002). We quantified a significant increase in

the amount of total VG as bees aged in the fall (p<0.0001), from very small amounts

upon emergence to an overall mean of ~60 цg/bee. There was also a significant

colony effect (p<0.0001), with the winter bees in late November in Colony 77 having

much more VG (~75 цg) than bees in Colony 91 (~30 цg). The age*colony

interaction was also highly significant (p<0.0001). There were very large differences

in the amounts of VG per bee. For example, in bees sampled at the end of

November amounts ranged from 0-220 цg/bee.

PUTATIVE STORAGE PROTEIN

This protein was found only in abdomens and is likely stored in the fat body. The

molecular size was indicative of other hexameric storage proteins in insects, and is

probably an arylphorin protein as that is the only class of storage protein that has

been identified in adult Apis (Danty et al., 1998). It will be referred to as arylphorin

subsequently.

The patterns in the amounts of arylphorin are almost identical to those discussed for

VG. There were highly significant effects of age (p<0.0001), colony (p<0.0001), and

age*colony interactions (p=0.015). Amounts of arylphorin were generally low at

emergence (~20 цg/bee), and increased gradually with age to wintering levels of ~75

цg/bee (range from 0-195 цg/bee).

THORAX PROTEIN (MUSCLE)

We are unsure what this protein is, but its abundance in the thorax suggests it is

related to thoracic musculature. This putative muscle protein increased quickly with

age to a maximum at around age 35 days, after which it appeared to decline slightly

with further increases in age.

PUTATIVE JELLY PROTEIN

This protein was detected in bee heads, suggesting it may have been a jelly protein.

Amounts were relatively small, with the amounts in bees from Colony 77 (41.0

цg/bee) and Colony 91 (25.7 цg/bee) being significantly different (p=0.003). There

was no significant effect of bee age on the amounts of this protein, although it

appeared that the amount of this protein in bees in colony 91 gradually increased

until they matched the levels found in bees from colony 77.

Page 4 of 6

APIACATA 38 (2004)352-357 355

DISCUSSION

It has been shown recently that adult worker honey bees contain a hexameric

storage protein, Hex70a, but details relating to its rate of accumulation were lacking

(Danty et al. 1998). In the absence of more detailed information about hexameric

storage proteins, research attention has focussed on vitellogenin (VG) which is the

most abundant haemolymph protein and a good indicator of the protein status of the

bee (Cremonez et al. 1998; Amdan and Omholt 2002). In our study of wintering

bees, we found a putative storage protein (probably an arylphorin) that increased

gradually over the fall, had the same pattern of accumulation as VG, and exceeded

the total amount of VG by approximately 25%. If VG is a true storage protein, as

concluded my Amdan and Omholt (2002), its amount in bees in early winter is

exceeded by quantities of arylphorin. The other proteins quantified in this study

(lipophorin, putative muscle protein, putative jelly protein) did not increase with bee

age, suggesting that they are not involved in protein storage to the same extent.

Both VG and arylphorin exhibited very strong colony effects, indicating that the

nutritive state of colonies within the same apiary can differ tremendously. In addition

to the colony effects on amounts of VG and arylphorin, additional colony effects were

detected in the amounts of lipophorin and putative jelly protein. Despite these strong

differences between bees from the two colonies, there were no outward signs during

bee collection to suggest the colonies differed in their protein nutrition. Poor protein

nutrition cannot be observed directly except by assessing amounts of stored pollen

and even that is problematic as it depends on many additional factors. Protein

nutrition undoubtedly affects colony health quite considerably, but goes largely

undiagnosed by beekeepers and bee scientists alike. The very large variability in

amounts of VG and arylphorin within the bees of a colony also raises questions as to

the past and current behavioural roles of bees by early winter that differ so greatly in

protein status.

Even summing the amounts of VG, arylphorin, and putative jelly protein in “well- endowed” winter bees at the end of November, the amount of stored protein is still

relatively small (e.g., ~400 цg). It seems that proteins stored within the bodies of

wintering bees are insufficient to maintain brood rearing for long. To put this into

perspective, Alfonsus (1933; cited by Ribbands 1953) calculated that the amount of

protein required to rear a single bee was ~29 mg. From our study, the amount stored

within an individual bee’s tissues that can be devoted to feeding larvae is nearly an

order of magnitude smaller. Our initial idea was that internally stored protein would

enable colonies to continue brood rearing in spring when pollen stores were depleted

and ambient conditions prohibited the collection of pollen from flowers, but the

amount of protein stored within a bee’s tissues is clearly insufficient to achieve that

purpose.

Why then do worker bees synthesize large storage proteins when stored pollen is

available within their nest? There are several possible explanations. First, evidence

is accumulating that storage proteins are a regular feature of insect biology, in many

cases as a way to store proteins when amino acids are in abundance temporally

and/or spatially. The continued synthesis of storage proteins in an insect that does

not require them may reflect the evolutionary history of social honey bees from

solitary ancestors for which protein storage may have been more important. A

second possibility may relate to the fact that over time the quality of pollen stored in

Page 5 of 6

APIACATA 38 (2004)352-357 356

honey bee nests decreases. Converting the amino acids in stored pollen into high- quality, easily accessible storage proteins may enable the worker bee to always have

a ready source of high quality protein for the production of jelly fed to colony

members (Crailsheim, 1992). A third possibility may relate to the early brood rearing

by workers while it is still winter. During cold snaps, contraction of the bee cluster

may prevent individual bees from accessing the pollen required for feeding larvae.

Collectively, workers with good internal supplies of stored proteins could continue to

feed some larvae during such periods of time, thereby buffering the effects of the

cold temperatures. For colonies infested with parasitic mites, good quantities of

storage proteins may enable bees to tolerate the negative effects of haemolymph

loss to mites. Finally, good protein nutrition may enable good functioning of the bee’s

immune system, enabling it to resist microbial infections both in the absence and

presence of parasitic mites. These explanations are mutually compatible and

demand attention from the research community.

ACKNOWLEDGMENTS

Paul Kelly maintains the bee hives at the University of Guelph. At the time of this

research, G. Otis was receiving substantial infrastructure support from the Ontario

Ministry of Agriculture and Food.

REFERENCES

Alfonsus A. (1933) Zum Pollenverbrauch des Bienenvolkes, Arch. Bienenk. 14, 220-

223.

Amdan AG, Omholt SW (2002) The regulatory anatomy of honeybee lifespan, J.

Theor. Boil. 216, 209-228.

Burmester T (1999) Evolution and function of the insect hexamerins. Eur. J. Entomol.

96, 213-225.

Crailsheim K (1992) The flow of jelly within a honeybee colony. J. Comp. Physiology

B. 162, 681-689.

Cremonez TA, de Jong D, Bitondi MM (1998) Quantification of hemolymph proteins

as a fast method for testing protein diets for honeybees. J. Econ. Entomol. 91, 1284-

1289.

Danty E, Arnold G, Burmester T, Huet JC, Huet D, Pernollet JC, Masson C (1998)

Identification and developmental profiles of hexamerins in the antenna and

hemolymph of the honeybee, Apis mellifera. Insect Biochem. Molec. Biol. 28, 387-

397.

RibbandsR (1953) The Behaviour and Social Life of Honeybees, Dover, New York

Seeley TD (1985) The Ecology of the Honey Bee, Princeton University, Princeton.

Wheeler DE, Buck NA (1995) Storage proteins in ants during development and

colony founding, J. Insect Physiology 41, 885-894.

Page 6 of 6

APIACATA 38 (2004)352-357 357

Wheeler DE, Martinez T (1994) Storage proteins in adult ants (Camponotus

festinates) – roles in colony founding by queens and in larval rearing by workers. J.

Insect Physiol. 40, 723-729.

Wheeler DE, Martinez T (1995) Storage proteins in ants (Hymenoptera, Formicidae).

Comp. Biochem. and Physiol. B. 112, 15-19.